PART1- Interview questionnaires:
Read the “Test Yourself” section on p. 199 in Ch. 9 of Exploring Research.
Discuss your response with your classmates.
Read the “Test Yourself” section on p. 207 in Ch. 9 of Exploring Psychology.
Discuss your response with your classmates.
PART3- LITERATURE REVIEW: Summarize what you have learned about the Literature Review process, this week.
Respond to one or more of the following prompts in one to two paragraphs
1. Provide citation and reference to the material(s) you discuss. Describe what you found interesting regarding this topic, and why.
2. Describe how you will apply that learning in your daily life, including your work life.
3. Describe what may be unclear to you, and what you would like to learn.
PART5-STEPS FOR CREATING METHODOLOGY: Using Figure 1.2 in Ch. 1 of Exploring Research, create a flowchart using Microsoft® Word or a similar program that helps you identify what research design to use for your research question.
RESOURCES FOR ASSIGNMENT 1 THROUGH 3:
Chapter 9 Nonexperimental Research: Descriptive and Correlational Methods
WHAT YOU’LL LEARN ABOUT IN THIS CHAPTER:
- • What survey research is as well as some of its advantages and disadvantages
- • The development and use of surveys
- • The value and validity of survey research
- • The importance and use of follow-up studies
- • The purpose and use of correlational research
- • How correlational studies are used
- • How to compute and interpret a correlation coefficient
In some ways, your work on the first eight chapters of Exploring Research has been done to prepare you for the next four, all of which deal with particular types of research designs or research methods. In this chapter, you will learn about nonexperimental research methods, which are ways of looking at research questions without the direct manipulation of a variable. Chapter 10 discusses another nonexperimental approach: qualitative methods. Why a separate chapter? Because the whole area of qualitative methods stands alone as a somewhat unique approach to asking and answering social and behavioral science research questions.So, let’s turn our attention to the techniques we will deal with here.For example, if you wanted to understand the factors that may be related to why certain undergraduates smoke and why others do not, you might want to complete some type of survey, one of the descriptive techniques that will be covered in this chapter. Or, if you were interested in better understanding the relationship between risk-taking behavior and drug abuse, perhaps the first (but not the last) step would be to conduct a correlational study in which you would learn about questions of a correlational nature. You would be examining the association between variables and learning about the important distinction between association (two things being related since they share something in common) and causality (one thing causing another).This chapter focuses on descriptive research questions, how they are asked and how they are answered. It’s the first chapter on methods before we move on to qualitative, true experimental, and quasi-experimental methods.
Although several factors distinguish different types of research from one another, probably the most important factor is the type of question that you want to answer (see the summary chart on page 00 in Chapter 1). If you are conducting descriptive research, you are trying to understand events that are occurring in the present and how they might relate to other factors. You generate questions and hypotheses, collect data, and continue as if you were conducting any type of research.Descriptive research describes the current state of some phenomenon.The purpose of descriptive research is to describe the current state of affairs at the time of the study. For example, if you want to know how many teachers use a particular teaching method, you could ask a group of students to complete a questionnaire, thereby measuring the outcome as it occurs. If you wanted to know whether there were differences in the frequency of use of particular types of words among 3-, 5-, and 7-year-olds, you would describe those differences within a descriptive or developmental framework.The most significant difference between descriptive research and causal comparative or experimental research (discussed in detail in Chapter 11) is that descriptive research does not include a treatment or a control group. You are not trying to test the influence of any variable upon another. In other words, all you are doing for readers of your research is painting a picture. When people read a report that includes one of the several descriptive methods that will be discussed, they should be able to envision the larger picture of what occurred. There may be room to discuss why it occurred, but that question is usually left to a more experimental approach.Although there are many different types of descriptive research, the focus of this discussion will be on survey research, and correlational studies in which relationships between variables are described.
The best application of sampling in theory and practice can probably be found in survey research. Survey researchers attempt to study directly the characteristics of populations through the use of surveys. You may be most familiar with the types of surveys done around election time, wherein relatively small samples of potential voters (about 1,200) are questioned about their voting intentions. To the credit of the survey designers, the results are often very close to the actual outcomes following the election.Survey research, also called sample surveys, examines the frequency and relationships between psychological and sociological variables and taps into constructs such as attitudes, beliefs, prejudices, preferences, and opinions. For example, a sample survey could be used to assess the following:
- • Parents’ attitudes toward the use of punishment in schools
- • Voting preferences
- • Neighborhood residents’ attitudes toward new parking restrictions
- • Adolescents’ perceptions of curfew enforcement
- • Use of drugs in high schools
- • A legislator’s views on capital punishment
The basic tool used in survey research is the interview. Interviews (or oral questionnaires) can take the form of the most informal question-and-answer session on the street to a highly structured, detailed interaction between interviewer and interviewee. In fact, many of the points that were listed for questionnaires also apply to interviews. For example, although you need not be concerned about the physical format of the questions in an interview (because the respondent never sees them), you do need to address such issues as transitioning between sections, being sensitive to the type of information you are requesting, and being objective and straightforward.Interviews are much more challenging and difficult to do well than just discussing a topic with someone.Most interviews begin with what is called face-sheet information, or neutral information, about the respondent such as age, living arrangements, number of children, income, gender, and educational level. Such information helps the interviewer accomplish several things.First, it helps establish a rapport between the interviewer and the interviewee. Such questions as “Where did you go to college?” or “How many children do you have?” are relatively nonthreatening.Second, it establishes a set of data that characterizes the person being interviewed. These data can prove invaluable in the analysis of the main focus of the interview which comes later on in the survey.Interviews contain two general types of questions: structured and unstructured questions. Structured or closed-ended questions have a clear and apparent focus and call for an explicit answer. They are comprehensible to the interviewer as to the interviewee. Such questions as “At what age did you start smoking?” and “How many times have you visited this store?” call for explicit answers. On the other hand, unstructured or open-ended questions allow the interviewee to elaborate upon responses. Such questions as “Why were you opposed to the first Persian Gulf War?” or “How would you address the issue of teenage pregnancy?” allow for a more broad response by the interviewee. In both cases, the interviewer can follow up with additional questions.Interviews can be especially helpful if you want to obtain information that might otherwise be difficult to come by, including firsthand knowledge of people’s feelings and perceptions. For example, in a study conducted by M. L. Smith and L. A. Shepard (1988), interviews with teachers and parents were part of a multifaceted approach to understanding kindergarten readiness and retention. In this study, interviewing was combined with other techniques such as in-class observations and the analysis of important documents. These researchers put the interview results to good use when they examined these outcomes in light of other information they collected throughout the study.On the positive side, interviews offer great flexibility by letting you pursue any direction (within the scope of the project) with the questions. You could also note the interviewee’s nonverbal behavior, the setting, and other information that might provide valuable information. Another advantage of interviews is that you can set the general tone and agenda at your own convenience (to a point, of course).There is also a downside to interviews. They take time, and time is expensive. Interviewing 10 people could take 20–30 hours including travel time and such. Also, because interviews have less anonymity than, for example, a questionnaire, respondents might be reluctant to come forward as honestly as they might otherwise. Other disadvantages are your own biases and the lack of a standardized set of questions. A good interviewer will probe deeply for additional information, perhaps of a different type, than would another interviewer who started with the same questions. Asking follow-up questions is an excellent practice, but what do you do about the interview where probing did not lead to the same information and thus produced different results?
What do you think a primary advantage of an interview is over a more structured tool such as a questionnaire, and when might you want to use the interview technique?
Developing an Interview
The development of an interview begins much like that for any proposal for a research project. Your first step is to state the purpose of the interview by taking into account your goals for the project. Then, as before, you review the relevant literature to find out what has been done in the past and whether other interview studies have been conducted. You may even find an actual interview that was previously used and be able to use parts of that in your own research. This is a very common practice when researchers use the same interview, say, 10 years later to look for changes in trends.Second, select a sample that is appropriate for your study, both in characteristics and in size. If you want to know about feelings regarding racial unrest, you cannot question only white citizens—you need to address all minorities. Similarly, even if interviews take lots of time and effort, you cannot skimp on sample size with the thought that what is lost in sample size can be made up in richness and detail. It does not work that way.Next, the interview questions need to be developed. As you know by now, questions, whether structured or unstructured, need to be clear and concise without any hidden agenda, double negatives, 75-cent words that cannot be understood, and so forth. One of the best ways to determine the appropriateness of your interview is by field-testing it. Use it with people who have the same characteristics as the intended audience. Listen to their feedback and make whatever changes you find necessary.After the interview form is (more or less) finished, it is time to train the interviewers. Most of the traits you want in an interviewer are obvious: They should be polite, neatly dressed, uncontroversial in appearance, and responsible enough to get to the interview site on time. These qualities, however, are not enough. Interviewers must learn how to go beyond the question should the need arise. For example, if you are asking questions about racial discrimination, the respondent might mention, “Yes, I sometimes feel as if I am being discriminated against.” For you not to ask “Why?” and to follow up on the respondent’s answer would result in the loss of potentially valuable and interesting information. The best way to train is to have an experienced interviewer watch the trainees interview a practice respondent and then provide feedback.Finally, it is time to conduct the actual interviews. Allow plenty of time, and go to it. Do not be shy, but do not be too aggressive either.
The Ten Commandments of Interviewing
If you have worked hard at getting ready for the interview, you should not encounter any major problems. Nonetheless, there are certain things you should keep in mind to make your interview run a bit more smoothly and be more useful later, when it comes time to examine the results of your efforts.No one is perfect, but you should strive to adhere to these 10 guidelines about interviewing as well as you can.With that in mind, here are the 10 commandments of interviewing (drumroll, please). Keep in mind that many, if not all of these, could also be classified as interviewer effects, in which the behavior of the interviewer can significantly affect the outcome.
- 1. Do not begin the interview cold. Warm up with some conversation about everything from the weather to the World Series (especially if there is a game that night and you know that the interviewee is a fan). Use anything you can to break the ice and warm up the interaction. If you are offered coffee, accept (and then do not drink all of it if you don’t want to). If you do not like coffee, politely refuse or ask for a substitute.
- 2. Remember that you are there to get information. Stay on task and use a printed set of questions to help you.
- 3. Be direct. Know your questions well enough so that you do not have to refer constantly to your sheet, but do not give the appearance that you are being too casual or uninterested.
- 4. Dress appropriately. Remove five of your six earrings if you feel wearing six would put off respondents. No shorts, no shirt, no interview, got it?
- 5. Find a quiet place where you and the interviewee will not be distracted. When you make the appointment for the interview, decide where this place will be. If a proposed location is not acceptable (such as “in the snack bar”), then suggest another (such as the lounge in the library). Call the day before your interview to confirm your visit. You will be amazed at how many interviewees forget.
- 6. If your interviewee does not give you a satisfactory answer the first time you ask a question, rephrase it. Continue to rephrase it in part or in whole until you get closer and closer to what you believe you need.
- 7. If possible, use a tape or digital recorder. If you do, you should be aware of several things. First, ask permission to tape the session before you begin. Second, the tape recorder should not be used as a crutch. Do not let the tape run without your taking notes and getting all the information you can while the interview is underway.
- 8. Make the interviewee feel like an important part of an important project, not just someone who is taking a test. Most people like to talk about things if given the chance. Tell interviewees you recognize how valuable their time is and how much you appreciate their participation. Be sure to promise them a copy of the results!
- 9. You become a good interviewer the same way you get to Carnegie Hall: practice, practice, practice.Your first interview, like everyone else’s, can be full of apprehension and doubt. As you do more of these, your increased confidence and mastery of the questions will produce a smoother process which will result in more useful information.
- 10. Thank the interviewee and ask if he or she has any questions. Offer to send (or call) the interviewee a summary of the results of your work.
Other Types of Surveys
Have you ever been at home during the dinner hour and the phone rings, and the person on the other end of the line wants to know how often you ride the bus, recycle your newspaper, use a computer, or rent a car?Those calls represent one of several types of survey research, all of which are descriptive in nature. In addition to interviews—the primary survey research method—and telephone surveys, surveys include panels or focus groups (in which a small group of respondents is interviewed and reinterviewed) and mail questionnaires.
How to Conduct Survey Research
Survey research starts out with a general plan (a flow plan) of what activities will occur when. The plan begins with the objective of the study, leads into the various methods that may be used to collect the data, and finishes with a final report and a summary of the findings.
- 1. Clarifying the objectives. The first step is to clarify the objectives of the survey research. For example, let’s say that a researcher is asked by a small school system to study attitudes toward the use of punishment in public schools. As part of the research plan, the researcher needs to consider the nature of the question being asked. Is the concern over the effectiveness of punishment? The way punishment is administered? The type of punishment (physical or other)?Defining the nature of the objectives may require some preliminary interviewing of respondents who might be interviewed in depth later in the project. One of the primary goals in this step of the project is to define the variables, such as punishment and attitudes, which are to be studied. Both of these terms, which are fairly vague by themselves, need further clarification and definition if the questions that are eventually asked by the researcher are to yield information of any importance.
- 2. Identifying a sample. After the objectives have been specified, the next step is to define a sampling plan and obtain a sample of individuals who will participate in the study. Will all teachers and parents be included? Probably not, because they would be too large a sample, and it would be inefficient to survey such a large group. But how can one fairly represent the community?Back to Chapter 4—how about taking a stratified random sample of three parents from each grade from four schools in the district, and a random sample of administrators from each of two administrative levels, building and central administration? If children are involved, the researcher may want to devise a plan that takes into account how frequently these children have been punished themselves and for what reason. Including only children who are rarely punished or only children who are always punished would skew the characteristics of the sample and, thus, the results.
- 3. Defining a method. Now that the objectives and the sampling plan are clear, exactly what will happen during the interview or panel study? Here are some of the questions about which a researcher may be concerned:
- • Will the questions be primarily open-ended, closed-ended, or a combination of both? How will each question sample content, opinions, or attitudes?
- • How will the sample of respondents be defined? Will it include parents, teachers, administrators, or all three? What about students?
- • How will the data be collected? Will interviews be used? Mail surveys?
- • What types of questions will be asked? What factual information will be included?
- These questions will be answered, in part, by the types of information the researcher needs to meet the objectives that were defined early in the project.
- 4. Coding and scoring. Survey research can result in anything from lengthy responses that have to be analyzed to a simple yes–no response, depending on the format and the content of the question. After the data have been collected, the researcher needs to code them (1 for male; 2 for female, for example) and then score the responses in an organized fashion that lends itself to easy tabulation.A simple example is shown in Table 9.1, which shows a breakdown of parents who regularly use physical punishment and those who do not and the judgments of both groups as to effectiveness of physical punishment.
Some type of analysis of the frequencies of these responses can be performed to answer the question about parents’ attitudes toward punishment.
The Validity of Survey Data
Collecting survey data is hard work. It means constantly seeking subjects and dealing with lots of extraneous sources of variance that are difficult to control. It is somewhat of a surprise, however, how relatively easy it is to establish the validity of such data. For example, one way to establish the validity of the data gained from an interview is to seek an alternative source for confirmation. Public records are easy to check to confirm such facts as age and party affiliation. Respondents can even be interviewed again to confirm the veracity of what they said the first time. There is no reason why people could not lie twice, but a good researcher is aware of that possibility and tries to confirm factual information that might be important to the study’s purpose.
Table 9.1 An example of how data can be collected and scored in a survey setting
Physical Punishment Is Cruel and IneffectivePhysical Punishment Is Harsh and UnnecessaryPhysical Punishment Can Work Under Certain ConditionsPhysical Punishment Is a Useful Deterrent for Poor BehaviorPhysical Punishment Is the Most Effective Method for Dealing with Poor BehaviorParents Who Use Punishment1214152332Parents Who Don’t Use Punishment461314 7 6
Evaluating Survey Research
Like all other research methods, survey research has its ups and downs. Here are some ups. First, survey research allows the researcher to get a very broad picture of whatever is being studied. If sampling is done properly, it is not hard to generalize to millions of people, as is done on a regular basis with campaign polling and such. Along with such powers to generalize comes a big savings in money and time.Second, survey research is efficient in that the data collection part of the study is finished after one contact is made with respondents and the information is collected. Also, minimal facilities are required. In some cases, just a clipboard and a questionnaire is enough to collect data.Third, if done properly and with minimal sampling error, surveys can yield remarkably accurate results.The downs can be serious. Most important are sources of bias which can arise during interviews and questionnaires. Interviewer bias occurs when the interviewer subtly biases the respondent to respond in one way or another. This bias might take place, for example, if the interviewer encourages (even in the most inadvertent fashion) approval or disapproval of a response by a smile, a frown, looking away, or some other action.On the other hand, the interviewee might respond with a bias because he or she may not want to give anything other than socially acceptable responses. After all, how many people would respond with a definite “yes!” to the question, “Do you beat your spouse?”These threats of bias must be guarded against by carefully training interviewers to be objective and by ensuring that the questions neither lead nor put respondents in a position where few alternatives are open.Another problem with survey research is that people may not respond, as in the case of a mail survey. Is this a big deal? It sure can be. Nonresponders might constitute a qualitatively distinct group from responders. Therefore, findings based on nonresponders will be different than if the entire group had been considered. The rule? Go back and try to get those who didn’t respond the first time.
You read about ethics and some guidelines in Chapter 3B. What might be some conflicts that can arise with those ethical principles and the use of the various survey methods we discussed earlier?
Correlational research describes the linear relationship between two or more variables without any hint of attributing the effect of one variable on another. As a descriptive technique, it is very powerful because this method indicates whether variables (such as number of hours of studying and test score) share something in common with each other. If they do, the two are correlated (or co-related) with one another.In Chapter 5, the correlation coefficient was used to estimate the reliability of a test. The same statistic is used here, again in a descriptive role. For example, correlations are used as the standard measure to assess the relationship between degree of family relatedness (e.g., twins, cousins, unrelated) and similarity of intelligence test scores. The higher the correlation, the higher the degree of relatedness. In such a case, you would expect that twins who are raised in the same home would have more similar IQ scores (they share more in common) than twins raised in different homes. And they do! Twins reared apart share only the same genetic endowment, whereas twins (whether monozygotic [one egg] or dizygotic [two eggs]) reared in the same home share both hereditary and environmental backgrounds.
The Relationship Between Variables
The most frequent measure used to assess degree of relatedness is the correlation coefficient, which is a numerical index that reflects the relationship between two variables. It is expressed as a number between 21.00 and 11.00, and it increases in strength as the amount of variance that one variable shares with another increases. That is, the more two things have in common (like identical twins), the more strongly related they will be to each other (which only makes sense). If you share common interests with someone, it is more likely that your activities will be related than if you compared yourself with someone with whom you have nothing in common.For example, you are more likely to find a stronger relationship between scores on a manual dexterity test and a test of eye–hand coordination than between a manual dexterity test and a person’s height. Similarly, you would expect the correlation between reading and mathematics scores to be stronger than that between reading and physical strength. This is because performances on reading and math tests share something in common with each other (intellectual and problem-solving skills, for example) than a reading test and, say, weight-lifting performance.Correlations can be direct or positive, meaning that as one variable changes in value, the other changes in the same direction, such as the relationship between the number of hours you study and your grade on an exam. Generally, the more you study, the better your grade will be. Likewise, the less you study, the worse your grade will be. Notice that the word “positive” is sometimes interpreted as being synonymous with “good.” Not so here. For example, there is a negative correlation between the amount of time parents spend with their children and the child’s level of involvement with juvenile authorities. Bad? Not at all.Positive correlations are not “good” and negative ones are not “bad.” Positive and negative have to do with the direction of the relationship and nothing else.Correlations can also reflect an indirect or negative relationship, meaning that as one variable changes in value in one direction, the other changes in the opposite direction, such as the relationship between the speed at which you go through multiple-choice items and your score on the test. Generally, the faster you go, the lower your score; the slower you go, the higher your score. Do not interpret this to mean that if you slow down, you will be smarter. Things do not work like that, which further exemplifies why correlations are not causal. What it means is that, for a specific set of students, there is a negative correlation between test-taking time and total score. Because it is a group statistic, it is difficult to conclude anything about individual performance and impossible to attribute causality.The two types of correlations we just discussed are summarized in Table 9.2.Interestingly, the important quality of a correlation coefficient is not its sign, but its absolute value. A correlation of 2.78 is stronger than a correlation of 1.68, just as a correlation of 1.56 is weaker than a correlation of 2.60.
What Correlation Coefficients Look Like
The most frequently used measure of relationships is the Pearson product moment correlation, represented by letter r followed by symbols representing the variables being correlated. The symbol rxyrepresents a correlation between the variables X and Y.To compute a correlation, you must have a pair of scores (such as a reading score and a math score) for each subject in the group with which you are working. For example, if you want to compute the correlation between the number of hours spent studying and test score, then you need to have a measure of the number of hours spent and a test score for each individual.The absolute value of the correlation coefficient, not the sign, is what’s important.As you just read, correlations can range between −1.00 and +1.00 and can take on any value between those two extremes. For example, look at Figure 9.1, which shows four sets of data (A, B, C, and D) represented by an accompanying scattergram for each of the sets.
Table 9.2 Two types of correlations: positive or direct, and negative or indirect
If X . . .and Y . . .The Correlation IsExampleIncreases in valueIncreases in valuePositive or directThe taller one gets (X), the more one weighs (Y)Decreases in valueDecreases in valuePositive or directThe fewer mistakes one makes (X), the fewer hours of remedial work (Y) one participates inIncreases in valueDecreases in valueNegative or indirectThe better one behaves (X), the fewer in-class suspensions (Y) one hasDecreases in valueIncreases in valueNegative or indirectThe less time one spends studying (X), the more errors one makes on the test (Y)The scattergram is a visual representation of the correlation coefficient of the relationship between two variables.A scattergram is a plot of the scores in pairs. In set A, the correlation is +1.70. (You will see how to compute that value in a moment.)To draw a scattergram, follow these steps:
- 1. Using graph paper, set up an X-axis (horizontal) and a Y-axis (vertical).
- 2. Indicate which variable from the pair will be X and which will be Y. The first in a pair is usually designated as the X value.
- 3. For participant 1, enter the coordinates for the X and Y values. In this example (data set A in Figure 9.1), the X score is 3 and the Y score is 3, so a data point corresponding to (3, 3) was entered.
- 4. Repeat step 3 for all the data points, and you will see the scattergram as shown in Figure 9.1 for data set A.
Now look at data set B, where the correlation is only .32, which is substantially weaker than .70. You can see that the stronger correlation (set A) is characterized in the following ways:
- • The data points group themselves closer and closer along a straight line as the correlation increases in strength.
- • As the slope of this grouping approaches a 45° angle, the correlation becomes stronger.
The data in set A show a high positive correlation (.70), whereas the data in set B show a much lower one (.32). The data in set C show a high negative correlation (− .82) and, just as with a high positive correlation, the coordinates that represent the intersection of two data points align themselves along a diagonal (in this case, from the upper left-hand corner to the lower right, approaching a 45° angle). The last data set, set D, shows very little relationship (− .15) between the X and the Y variables, and the accompanying plot of the coordinates reveals a weak pattern. In other words, a line drawn through these points would be almost flat or horizontal.
In summary, the stronger the formation of a pattern and the more the pattern aligns itself in a 45° angle (either from the lower left-hand corner of the graph to the upper right-hand for positive correlations, or from the upper left-hand corner of the graph to the lower right-hand corner for negative correlations), the stronger the visual evidence of the existence of a relationship between two variables.
Correlations can be negative or positive, but give an example of how negative does not carry a pejorative meaning and positive outcomes are not always good.
Computing the Pearson Correlation Coefficient
The easiest manual way to compute the correlation between two variables is through the use of the raw score method. The formula for rxy (where xy represents the correlation between X and Y) is as follows:The Pearson correlation coefficient is the most frequently computed type of correlation.r xy = n Σ XY − Σ X Σ Y√ [ n Σ X 2 − ( Σ X ) 2 ] [ n Σ Y 2 − ( Σ Y ) 2 ] where
- r xy = the correlation coefficient between X and Y Σ = the summation sign n = the size of the sample X = the individual’s score on the X variable Y = the individual’s score on the Y variable XY = the product of each X score times its corresponding Y score X 2 = the individual X score, squared Y 2 = the individual Y score, squared
Let’s look at a simple example where the correlation coefficient is computed from data set C shown in Figure 9.1. The mean for variable X is 6.3, and the mean for variable Y is 4.6. Here is what the finished equation looks like:r xy = − .373 [ 32.1 ] [ 62.4 ] = − .82Try it yourself and see if you can get the same result (rxy = −.82). You can also use SPSS or Excel to get the answer.
Table 9.3 An example of more than two variables and the possible correlations between them
GradeReadingMathGrade1.00.321.039Reading.3211.00.605Math.039.6051.00The correlation is the expression of the relationship between the variables of X and Y, represented as rxy. What happens if you have more than two variables? Then you have more than one correlation coefficient. In general, if you have n variables, then you will have “n taken two at a time” pairs of relationships. In Table 9.3, you can see a correlation matrix, or a table revealing the pairwise correlations between three variables (grade, reading score, and mathematics score). Each of the three correlation coefficients was computed by using the formula described earlier.You may notice that the diagonal of the matrix is filled with 1.00s because the correlation of anything with itself is always 1. Also, the coefficients to the right of the diagonal and to its left form a mirror image. The correlations for the other “half” of the matrix (above or below the diagonal of 1.00s in Table 9.3) are the same.
Interpreting the Pearson Correlation Coefficient
The correlation coefficient is an interesting index. It reflects the degree of relationship between variables, but it is relatively difficult to interpret as it stands. However, there are two ways to interpret these general indicators of relationships.To interpret the meaning of the correlation coefficient, look to the correlation of determination.The first method is the “eyeball” method, in which correlations of a certain value are associated with a certain nominal degree of relationship such that:Correlations betweenAre said to be.8 and 1.0Very strong.6 and .8Strong.4 and .6Moderate.2 and .4Weak.0 and .2Very weakRemember: Do not be fooled by these numbers. Even the weakest correlation (such as .1) can be statistically significant if the sample upon which it is based is large enough and sufficiently approaches the size of the population. You read about the significance versus meaningfulness distinction in Chapter 8.
Table 9.4 Differences in the amount of variance accounted for as a function of different values of the correlation coefficient
If rxy isand rxy2 isthen the change from . . .accounts for this much more variance (%)0.10.01 0.20.04.1 to .230.30.09.2 to .3184.108.40.206 to .4220.127.116.11 to .518.104.22.168 to .622.214.171.124 to .7126.96.36.199 to .8188.8.131.52 to .9171.01.00.9 to 1.019A sounder method for interpreting the correlation coefficient is to square its value and then compute the coefficient of determination. This value, rxy2, is the amount of variance that is accounted for in one variable by the other. In other words, it allows you to estimate the amount of variance that can be accounted for in one variable by examining the amount of variance in another variable. Thus, if the correlation between two variables is .40, then the coefficient of determination is .16. Sixteen percent (16%) of the variance in one variable can be explained by the variance in the other variable; 84% (or 100%-16%) of the variance is unexplained. This portion of unexplained variance is often referred to as the coefficient of alienation.
It is interesting to compare how the amount of variance explained in the relationship between two variables changes as the correlation gets stronger. The change isn’t as predictable as you might think.Table 9.4 shows the simple correlation coefficient (the first column) and the coefficient of determination (the second column). Notice the change in the amount of variance accounted for as the value of the correlation coefficient increases. For example, if the correlation is increased from .4 to .5, the increase in the amount of variance accounted for is 9%. But if the correlation is increased a similar amount (say, from .6 to .7, which is still .1), then the increase in the amount of variance accounted for is 13%. The increase in the variance explained is not linear; therefore, the higher the correlation is, the larger the “jump” in explained variances.Figure 9.2 is a graphic illustration of what is shown in Table 9.4. As the correlation increases in value, an increasingly larger amount of variance is accounted for. That’s why the line shown in Figure 9.2 curves—the amount of variance (Y) increases disproportionately as the value of the correlation coefficient (the Xaxis) increases and that’s why the higher the value of the correlation, the more relative variance you can explain as a relationship between variance than for a lower correlation value.
Of the various research method tools you have learned about so far, what are some of the advantages and disadvantages of the correlational research methods?
Is a nonexperimental—descriptive or correlational—design right for you? This is not really the question that should be asked. Rather, you should ask if your subject of interest demands that you use the tools suggested by the descriptive method. As emphasized before, the question that is asked determines the way it is answered. If you want to investigate how the Oklahoma settlers of the 1930s raised their children or how child rearing has changed, historiography may be for you. And what does the descriptive method offer? It provides an account of an event, often in such detail that it serves as a springboard for other questions to be asked and answered. Case studies, developmental research, and correlational studies describe a particular phenomenon in a way that communicates the overall picture of whatever is being studied. Although these methods do not allow the luxury of implying any cause-and-effect relationship between variables, their use provides the tools needed to answer questions that are otherwise unanswerable.
1.Write out several questions that would be interesting to study using survey research. Create a few questions of a survey nature for each of the studies.2.Name two advantages and two disadvantages to interviews.3.Write three potential follow-up questions to this initial interview question: What is your attitude toward eliminating score keeping in children’s sports?4.Briefly outline the five steps of developing an interview.5.Rank the following correlation coefficients in order of their strength from strongest to weakest.
6.What is wrong with the following argument? The relationship between the number of hours you spend studying is directly related to how well you do on school tests. Therefore, if you do not do well on a test, it means that you did not study long enough.7.Indicate the type of correlation each of the following relationships describes: positive, negative, or no relationship.
- (a)As A increases, B increases in value.
- (b)As A increases, B decreases in value.
- (c)As A decreases, B increases in value.
8.For each of the three relationships in exercise #4, provide an example.9.Tell whether the following hypotheses are correlational in nature.
- (a)There are no differences in cognitive ability between preschoolers in child-care settings and preschoolers who are cared for at home.
- (b)There is a relationship between parents’ education, socioeconomic status, and children’s achievement levels in math.
- (c)There is no relationship between the rate of violent crime in New York and socioeconomic status.
- (d)Parent education does not increase a child’s performance on a math achievement test.
- (e)Over time, there are differences in the discipline policies used in rural and urban schools.
10.Improve the following interview questions:
- (a)Do you think it is acceptable for teenagers to smoke cigarettes and drink alcohol?
- (b)What are the most common colloquialisms used at your junior high school?
11.What is the purpose of descriptive research?12.Provide an example of when descriptive research might be the appropriate method to use to answer your research question. And while you are at it, what is your question?13.Which of the following statements about correlation coefficients are true?
- (a)Correlations can be positive.
- (b)Correlations can be negative.
- (c)Correlations reflect causation.
- (d)Correlations measure the relationship between two variables.
14.What is an example of where a correlation might be significant but not meaningful?15.Examine the relationship between consumption of milk during dinner and nighttime bedwetting and find a significant correlation of .25. How would you interpret the meaningfulness of this finding?16.What does the coefficient of determination mean? What would the value of the coefficient of determination be for two variables with a correlation of .60? What would be the value of the coefficient of alienation?
Online. . .
A huge number of educational databases (as part of the main reading room of the Library of Congress) to start your own descriptive research can be found at http://www.loc.gov/rr/main/alcove9/education/database.html. You can find everything here from ERIC to a listing of universities worldwide.
Using Interviews in Research
Dr. John Suler at Rider University gives tips on how to conduct the interview and how appropriately to include information from interviews in your research paper at http://www-usr.rider.edu/~suler/interviews.html.
RESOURCES FOR PART 4:
Chapter 1 The Role and Importance of Research
WHAT YOU’LL LEARN ABOUT IN THIS CHAPTER:
- • Who does research and why
- • How research is defined and what some of its purposes are
- • What a model of scientific inquiry is and how it guides research activities
- • Some of the things that research is and some of the things that it isn’t
- • What researchers do and how they do it
- • The characteristics of good research
- • How a method of scientific inquiry guides research activity
- • The different types of research methods and examples of each
Say Hello to Research!
Walk down the hall in any building on your campus where social and behavioral science professors have their offices in such departments as psychology, education, nursing, sociology, and human development. Do you see any bearded, disheveled, white-coated men wearing rumpled pants and smoking pipes, hunched over their computers and mumbling to themselves? How about disheveled, white-coated women wearing rumpled skirts, smoking pipes, hunched over their computers, and mumbling to themselves?Researchers hard at work? No. Stereotypes of what scientists look like and do? Yes. What you are more likely to see in the halls of your classroom building or in your adviser’s office are men and women of all ages who are hard at work. They are committed to finding the answer to just another piece of the great puzzle that helps us understand human behavior a little better than the previous generation of scientists.Like everyone else, these people go to work in the morning, but unlike many others, these researchers have a passion for understanding what they study and for coming as close as possible to finding the “truth.” Although these truths can be elusive and sometimes even unobtainable, researchers work toward discovering them for the satisfaction of answering important questions and then using this new information to help others. Early intervention programs, treatments of psychopathology, new curricula, conflict resolution techniques, effective drug treatment programs, and even changes in policy and law have resulted from evidence collected by researchers. Although not always perfect, each little bit of evidence gained from a new study or a new idea for a study contributes to a vast legacy of knowledge for the next generation of researchers such as yourself.You may already know and appreciate something about the world of research. The purpose of this book is to provide you with the tools you need to do even more, such as
- • develop an understanding of the research process.
- • prepare yourself to conduct research of your own.
- • learn how to judge the quality of research.
- • learn how to read, search through, and summarize other research.
- • learn the value of research activities conducted online.
- • reveal the mysteries of basic statistics and show you how easily they can be used.
Today, more than ever, decisions are evidence based, and what these researchers do is collect evidence that serves as a basis for informed decisions.
- • measure the behaviors, traits, or attributes that interest you.
- • collect the type of data that relate to your area of interest.
- • use a leading statistical package (SPSS) to analyze data.
- • design research studies that answer the question that you want answered.
- • write the type of research proposal (and a research report) that puts you in control—one that shows you have command of the content of the research as well as the way in which the research should be done.
Sound ambitious? A bit terrifying? Exciting? Maybe those and more, but boring is one thing this research endeavor is not. This statement is especially true when you consider that the work you might be doing in this class, as well as the research proposal that you might write, could hold the key to expanding our knowledge and understanding of human behavior and, indirectly, eventually helping others.So here you are, beginning what is probably your first course in the area of research methods and wondering about everything from what researchers do to what your topic will be for your thesis. Relax. Thousands of students have been here before you and almost all of them have left with a working knowledge of what research is, how it is done, and what distinguishes a good research project from one that is doomed. Hold on and let’s go. This trip will be exciting.
What Research Is and What It Isn’t
Perhaps it is best to begin by looking at what researchers really do. To do so, why not look at some of the best? Here are some researchers, the awards they have won, and the focus of their work. All of these people started out in a class just like the one you are in, reading a book similar to the one you are reading. Their interest in research and a particular issue continued to grow until it became their life’s work.The following awards were given in 2009 by the American Psychological Association in recognition of outstanding work.Research is, among other things, an intensive activity that is based on the work of others and generates new ideas to pursue and questions to answer.Susan E. Carey from the psychology department at Harvard University was honored for her contributions to the field of cognitive development and developmental psychology. The work that she did early in her career focused on understanding how children learn language, and she coined the term “fast mapping” for how children can learn the meaning of a new word with very little experience with that word.Nancy E. Adler from the University of California won the Distinguished Scientific Award for the Applications of Psychology for her work in health. Her early research focused on the health behaviors in adolescence, and she explained the incredibly interesting question of why individuals engage in health-damaging behaviors and how their understanding of risk affects their choices.Finally, one of several Distinguished Scientific Awards for Early Career Contributions to Psychology went to Jennifer A. Richeson from Northwestern University for her work on stereotyping, prejudice, discrimination, and inter-group conflict. This focus examined the experiences and behaviors both of members of devalued groups and of members of dominant groups.The American Educational Research Association (AERA) also gives out awards that recognize important contributions.The 2009 E. F. Lindquist award was given to Wim J. van der Linden for his contributions to the field of testing and measurement, including optimal test design and adaptive testing. The award is named after E. F. Lindquist, who was a founder of The American College Testing Program, and is given for outstanding applied or theoretical research in the field of testing and measurement.AERA has an extensive award program including the Distinguished Contributions to Gender Equity in Education Research Award, given to Sandra Harding from the University of California–Los Angeles in recognition of her research that helps to advance public understanding of gender and/or sexuality in the education community.And, as with many other organizations, AERA also offers awards for researchers still early in their careers, such as the Early Career Award won by Michele Moses from the University of Colorado–Boulder and Nell Duke from Michigan State University.What all these people have in common is that at one time or another during their professional careers, they were active participants in the process of doing research. Research is a process through which new knowledge is discovered. A theory, such as a theory of motivation, or development, or learning, for example, helps us to organize this new information into a coherent body, a set of related ideas that explain events that have occurred and predict events that may happen. Theories are an important part of science. It is at the ground-floor level, however, that the researcher works to get the ball rolling, adding a bit of new insight here and a new speculation there, until these factors come together to form a corpus of knowledge.High-quality research is characterized by many different attributes, many of which tend to be related to one another and also tend to overlap. High-quality research
- • is based on the work of others,
- • can be replicated,
- • is generalizable to other settings,
- • is based on some logical rationale and tied to theory,
- • is doable,
- • generates new questions or is cyclical in nature,
- • is incremental, and
- • is an apolitical activity that should be undertaken for the betterment of society.
Let’s take a closer look at each of these.First, research is an activity based on the work of others. No, this does not mean that you copy the work of others (that’s plagiarism), but you always look to the work that has already been done to provide a basis for the subject of your research and how you might conduct your own work. For example, if there have been 200 studies on gender differences in aggression, the results of those studies should not be ignored. You may not want to replicate any one of these studies, but you certainly should take methodologies that were used and the results into consideration when you plan your own research in that area.A good example of this principle is the tremendous intellectual and scientific effort that went into the creation of the atomic bomb. Hundreds of top scientists from all over the world were organized at different locations in an intense and highly charged effort to combine their knowledge to create this horrible weapon. What was unique about this effort is that it was compressed in time; many people who would probably share each other’s work in any case did so in days rather than months because of the military and political urgency of the times. What was discovered one day literally became the basis for the next day’s experiments (see Richard Rhodes’ Pulitzer Prize–winning book, The Making of the Atomic Bomb, for the whole story).Second, while we’re talking about other studies, research is an activity that can be replicated. If someone conducts a research study that examines the relationship between problem-solving ability and musical talent, then the methods and procedures (and results) of the experiment should be replicable with other groups for two reasons. First, one of the hallmarks of any credible scientific finding is that it can be replicated. If you can spin gold from straw, you should be able to do it every time, right? How about using a new method to teach children to read? Or developing early intervention programs that produce similar results when repeated? Second, if the results of an experiment can be replicated, they can serve as a basis for further research in the same area.Third, good research is generalizable to other settings. This means, for example, that if adolescent boys are found to be particularly susceptible to peer pressure in one setting, then the results would probably stand up (or be generalizable) in a different but related setting. Some research has limited generalizability because it is difficult to replicate the exact conditions under which the research was carried out, but the results of most research can lend at least something to another setting.Fourth, research is based on some logical rationale and tied to theory. Research ideas do not stand alone merely as interesting questions. Instead, research activity provides answers to questions that help fill in pieces to what can be a large and complicated puzzle. No one could be expected to understand, through one grand research project, the entire process of intellectual development in children, or the reason why adolescents form cliques, or what actually happens during a midlife crisis. All these major areas of research need to be broken into smaller elements, and all these elements need to be tied together with a common theme, which more often than not is some underlying, guiding theory.Fifth, and by all means, research is doable! Too often, especially for the young or inexperienced scientist (such as yourself), the challenge to come up with a feasible idea is so pressing that almost anything will do as a research topic. Professors sometimes see thesis statements from students such as, “The purpose of this research is to see if the use of drugs can be reduced through exposure to television commercials.” This level of ambiguity and lack of a conceptual framework makes the statement almost useless and certainly not doable. Good research poses a question that can be answered, and then answers it in a timely fashion.Sixth, research generates new questions or is cyclical in nature. Yes, what goes around comes around. The answers to today’s research questions provide the foundation for research questions that will be asked tomorrow. You will learn more about this process later in this chapter when a method of scientific inquiry is described.Seventh, research is incremental. No one scientist stands alone; instead, scientists stand on the shoulders of others. Contributions that are made usually take place in small, easily definable chunks. The first study ever done on the development of language did not answer all the questions about language acquisition, nor did the most recent study put the icing on the cake. Rather, all the studies in a particular area come together to produce a body of knowledge that is shared by different researchers and provides the basis for further research. The whole, or all the knowledge about a particular area, is more than the sum of the parts, because each new research advance not only informs us but it also helps us place other findings in a different, often fruitful perspective.Finally, at its best, research is an apolitical activity that should be undertaken for the betterment of society. I stress “at its best,” because too often various special-interest groups dictate how research funding should be spent. Finding a vaccine for acquired immunodeficiency syndrome (AIDS) should not depend on one’s attitudes toward individual lifestyles. Similarly, whether early intervention programs should be supported is independent of one’s personal or political views. And should research on cloning be abandoned because of its potential misuse? Of course not. It’s how the discovery of new knowledge is used that results in its misuse, not the new knowledge itself.Although it should be apolitical, research should have as its ultimate goal the betterment of society. Researchers or practitioners do not withhold food from pregnant women to study the effects of malnutrition on children. To examine the stress-nutrition link, researchers do not force adults to eat particular diets that might be unhealthy. These unethical practices would not lead to a greater end, especially because there are other ways to answer such questions without resorting to possibly harmful practices.If these attributes make for good research, what is bad research? It takes the opposite approach of all the things stated earlier and then some. In sum, bad research is the fishing trip you take looking for something important when it simply is not to be found. It is plagiarizing other people’s work, or falsifying data to prove a point, or misrepresenting information and misleading participants. Unfortunately, there are researchers whose work is characterized by these practices, but they are part of an overall minority.
Note: At the end of every major heading in each chapter of Exploring Research, we’ll have a few questions for you that we hope will help you understand the content and guide your studying.Provide an example of how research is incremental in nature and what advantage is this to both future and past researchers?Think of an example of how knowledge about a certain topic can lead to new questions about that, or a related, topic.
A Model of Scientific Inquiry
In the past 20 years, the public has been exposed to the trials and tribulations of the research process as described through hundreds of books by and about the everyday work of scientists around the world.Regardless of the specific content of these books, they all have one thing in common. The work was accomplished through adherence to guidelines that allowed these researchers to progress from point A to point Z while remaining confident that they were on the trail of finding (what they hoped was) an adequate answer to the questions they had posed.“Doing science” means following a model that begins with a question and ends with asking new questions.Their methods and their conclusions are not helter-skelter because of one important practice: They share the same general philosophy regarding how questions about human behavior should be answered. In addition, for scientists to be able to trust their colleagues, in the sense of having confidence in the results produced by their studies, these scientists must have something in common besides good intentions. As it turns out, what they share is a standard sequence of steps in formulating and answering a question.When you read in a journal article that Method A is more effective than Method B for improving retention or memory, you can be pretty sure that the steps described next were followed, in one form or another. Because there is agreement about the general method used to answer the question, the results of this comparison of Method A and Method B can be applied to the next study. That study would perhaps investigate variations of Method A and how and why they work. The research efforts of developmental psychologists, gerontologists (specialists in aging), linguists, and experts in higher education all depend on the integrity of the process.Figure 1.1 shows a set of such steps as part of a model of scientific inquiry. The goal of this model is to find the truth (whatever that means) or, in other words, to use a scientific method that results in a reasonable and sound answer to important questions that will further our understanding of human behavior.An interesting and timely topic, the effects of using social media on adolescents’ social skills, will be used as an example of the different steps followed in this model.
Asking the Question
Remember the story of The Wizard of Oz? When Dorothy realized her need to get to the Emerald City, she asked Glinda, the good witch, “But where do I begin?” Glinda’s response, “Most people begin at the beginning, my dear,” is the case in almost any scientific endeavor.Our first and most important step is asking a question (I wonder what would happen if . . . ?) or identifying a need (We have to find a way to . . .) that arises as the result of curiosity, and to which it becomes necessary to find an answer. For example, you might be curious about how the use of social media such as Twitter and Facebook affects relationships between children and their peers. You also might feel an urgency to find out how to use various types of media most effectively for educating children and adults about the dangers of using drugs.Such questions are informally stated and often are intended as a source of discussion and stimulation about what direction the specific research topic should take. Where do such questions come from? They rarely come from the confines of a classroom or a laboratory. Rather, questions spring (in the fullest sense of the word) from our imagination and our own experiences, enriched by the worlds of science, art, music, and literature. It is no coincidence that many works of fiction (including science fiction) have a basis in fact. The truly creative scientist is always thinking about everything from solutions to existing questions to the next important question to ask. When Louis Pasteur said that chance favors the prepared mind, he was really saying, “Take advantage of all the experiences you can, both in and out of school.” Only then can you be well prepared to recognize the importance of certain events, which will act as a stimulus for more rigorous research activity.Questions can be as broad as inquiring about the effects of social media on peer groups, or as specific as the relationship between the content of social media transactions and acceptance by peers. Whatever their content or depth of inquiry, questions are the first step in any scientific endeavor.
Identifying the Important Factors
Once the question has been asked, the next step is to identify the factors that have to be examined to answer the question. Such factors might range from the simplest, such as an adolescent’s age or socioeconomic status, to more complicated measures, such as the daily number of face-to-face interactions.For example, the following list of factors have been investigated over the past 10 years by various researchers who have been interested in the effects of social media:
- • age and gender of the adolescent,
- • ethnicity,
- • level of family education,
- • access to types of social media,
- • number of self-identified close friends,
- • parental attitude toward social media,
- • family configuration,
- • family communication patterns.
And these are only ten of hundreds of factors and associated topics that could be explored. But of all the factors that could be important and that could help us to understand more about the effects of social media, which ones should be selected as a focus?In general, you should select factors that
- • have not been investigated before,
- • will contribute to the understanding of the question you are asking,
- • are available to investigate,
- • hold some interest for you personally or professionally,
- • lead to another question.
It is hard enough to define the nature of the problem you want to study (see Chapter 3) let alone generate questions that lead to more questions, but once you begin the journey of becoming a scientist, you are a member of an elite group who has the responsibility to contribute to the scientific literature not only by what you do but also by what you see that needs to be done.
Formulating a Hypothesis
When asked what she thought a hypothesis was, a 9-year-old girl said it best: “An educated guess.” A hypothesis results when the questions are transformed into statements that express the relationships between variables such as an “if . . . then” statement.For example, if the question is, “What effects does using Facebook have on the development of friendships?” then the hypothesis could be, adolescents who use Facebook as their primary means of maintaining social contact have fewer close friends. Several characteristics make some hypotheses better than others, and we will talk about those in Chapter 2.For now, you should realize that a hypothesis is an objective extension of the question that was originally posed. Although all questions might not be answerable because of the way in which they are posed—which is fine for the question stage—a good hypothesis poses a question in a testable form. Good questions lead to good hypotheses, which in turn lead to good studies.
Collecting Relevant Information
Hypotheses should posit a clear relationship between different factors, such as a correlation between number of followers on Twitter and quality of social skills. That is the purpose of the hypothesis. Once a hypothesis is formulated, the next step is the collection of information or empirical data that will confirm or refute the hypothesis. So, if you are interested in whether or not participating in social media has an impact on adolescent’s social skills, the kinds of data that will allow the hypothesis to be tested must be collected.For example, you might collect two types of data to test the hypothesis mentioned in the previous paragraph. The first might be the number of friends an adolescent might have. The second might be the quality of those relationships.An important point about testing hypotheses is that you set out to test them, not to prove them. As a good scientist, you should be intent on collecting data that reveal as much of the truth about the world as is possible and letting the chips fall where they may, whether you agree or disagree with the outcomes. Setting out to prove a hypothesis can place scientists in the unattractive position of biasing the methods for collecting data or the way in which study results are interpreted. If bias occurs, then the entire sequence of steps can fall apart. Besides, there’s really no being “wrong” in science. Not having a hypothesis supported means only that there are additional questions to ask or that those which were asked should be reformulated. That is the beauty of good science—there is always another question to ask on the same topic—one that can shed just a bit more light. And who knows? That bit more light might be the tipping point or just the amount needed to uncover an entirely new and significant finding.
Testing the Hypothesis
Is it enough simply to collect data that relate to the phenomena being studied? Not quite. What if you have finished collecting data and find that adolescents who spend more than 10 hours a week involved in social media have 50% fewer qualitatively “good” relationships with peers than those who spend less than 10 hours? What would your conclusion be?On one hand, you could say the adolescents who used social media more than 10 hours per week were one-half as sociable as other adolescents or had one-half the quality of relationships of the children who used social media less than 10 hours per week. On the other hand, you might argue that the difference between the two groups of adolescents is too large enough for you to reach any conclusion. You might conclude that in order for a statement about social media use and quality of friendships, you would have to have much greater differences in the quality of relationships.Say hello to inferential statistics (see Chapter 8 for more), a set of tools that allows researchers to separate the effects of an isolated factor (such as time spent on Facebook) from differences between groups that might be owing to some other factor or to nothing other than chance. Yes, luck, fate, destiny, the wheels of fortune, or whatever you want to call what you cannot control, sometimes can be responsible for differences between groups.For example, what if some of the adolescents participating in your study went to some kind of social function where there was a particularly strong emphasis on social media methods of communicating such as texting. Or, what if one of the adolescents just was afraid to truthfully report how much time he or she spent on Facebook during study time?The job of all the tools that researchers have at their disposal (and the ones you will learn about throughout Exploring Research) is to help you separate the effects of the factors being studied (such as amount of time spent on Facebook) from other unrelated factors (such as the number of years a family has lived at its current address). What these tools allow researchers to do is assign a probability level to an outcome so that you can decide whether what you see is really due to what you think it is due to or something else which you leave for the next study.
Working with the Hypothesis
Once you have collected the required data and have tested the hypothesis, as a good scientist you can sit down, put up your feet, look intellectual, and examine the results. The results may confirm or refute the hypothesis. In either case, it is off to the races. If the data confirm your hypothesis, then the importance of the factors that were hypothesized to be related and conceptually important were borne out and you can go on your merry way while the next scientific experiment is being planned. If the hypothesis is not confirmed, it may very well be a time for learning something that was not known previously. In the example used earlier, it may mean that involvement in social media has no impact on social skills or social relationships. Although the researcher might be a bit disappointed that the initial hunch (formally called a hypothesis) was not supported, the results of a well-run study always provide valuable information, regardless of the outcome.
Reconsidering the Theory
Finally, it is time to take stock and relate all these research efforts to what guides our work in the first place: theory. Earlier in this chapter, a theory was defined as a set of statements that predict things that will occur in the future and explain things that have occurred in the past. But the very nature of theories is that they can be modified according to the results of research based on the same assumptions on which the theory is based.For example, a particular approach to understanding the development of children and adults is known as social learning theory, which places special importance on the role of modeling and vicarious, or indirect, learning. According to this theory, exposure to aggressive behavior would lead to aggressive behavior once the environment contains the same kinds of cues and motivation that were present when the initial aggressive model (such as particularly unkind Facebook postings) was observed.If the hypothesis that observing such models increases lack of civility is confirmed, then another building block, or piece of evidence, has been added to the house called social learning theory. Good scientists are always trying to see what type of brick (new information) fits where, or if it fits at all. In this way, new knowledge can change or modify the way the theory appears and what it has to say about human behavior. Consequently, new questions might be generated from the theory that will help contribute further to the way in which the house is structured.
Asking New Questions
In any case, the last step in this simple model of scientific inquiry is to ask a new question. It might be a simple variation on a theme (Do males use social media in a different way than females?) or a refinement of the original question (How might the use of social media differentially affect the social relationships of males and females?). Whether or not the hypothesis is supported, good research leaves you farther along the trail to answering the original question. You just might be at a different place than you thought or intended to be.
Hypothesis plays a very important role in scientific research, with one of them being the objective testing of a particular question that a scientist might want to ask. What are some of the factors that might get in the way of the scientist remaining objective and what impact might that have on a fair test of the hypothesis of interest? What is the danger of not being aware of these biases?
Different Types of Research
By now, you have a good idea what research is and how the research process works. Now it is time to turn your attention to a description and examples of different types of research methods and the type of questions posed by them.The types of research methods that will be discussed differ primarily on three dimensions: (1) the nature of the question asked, (2) the method used to answer it, and (3) the degree of precision the method brings to answering the question. One way in which these methods do not necessarily differ, however, is in the content or the focus of the research.In other words, if you are interested in the effects of the use of social media on adolescents’ friendships, your research may be experimental, where you artificially restrict access to social media and look at friendship outcomes, or nonexperimental, where you survey a group of adolescents to determine the frequency of use of social media tools.A summary of the two general categories of research methods (nonexperimental versus experimental), which will be discussed in this volume, is shown in Table 1.1. This table illustrates the purpose of each category, the time frame that each encompasses, the degree of control the different method has over competing factors, “code” words that appear in research articles that can tip you off as to the type of research being conducted, and an example of each. Chapters 9–12 discuss in greater detail each of these research methods.There is one very important point to keep in mind when discussing different methods used in research. As often as not, as research becomes more sophisticated and researchers (like you in the future) become better trained, there will be increased reliance on mixed methods models, where both experimental and nonexperimental methods are combined. Some researchers feel that this type of approach lacks clarity and precision, but others feel it is the best way to look at a phenomenon of interest from a variety of perspectives and thereby be more informative.
Nonexperimental research includes a variety of different methods that describe relationships between variables. The important distinction between nonexperimental methods and the others you will learn about later is that nonexperimental research methods do not set out, nor can they test, any causal relationships between variables. For example, if you wanted to survey the social media–using behavior of adolescents, you could do so by having them maintain a diary in which they record what tools they use and for how long.Nonexperimental research examines the relationship between variables, without any attention to cause-and-effect relationships.
Table 1.1 Summary of research methods covered in exploring research.
Types of ResearchNonexperimentalExperimental DescriptiveHistoricalCorrelationalQualitativeTrue ExperimentalQuasi-experimentalPurposeDescribe the characteristics of an existing phenomenonRelate events that have occurred in the past to current eventsExamine the relationships between variablesTo examine human behavior and the social, cultural, and political contexts within which it occursTo test for true cause-and-effect relationshipsTo test for causal relationships without having full controlTime frameCurrentPastCurrent or past (correlation) Future (prediction)Current or pastCurrentCurrent or pastDegree of control over factors or precisionNone or lowNone or lowLow to mediumModerate to highHighModerate to highCode words to look for in research articlesDescribe Interview Review literaturePast DescribeRelationship Related to Associated with PredictsCase study Evaluation Ethnography Historical Research SurveyFunction of Cause of Comparison Between Effects ofFunction of Cause of Comparison between Effects ofExampleA survey of dating practices of adolescent girlsAn analysis of Freud’s use of hypnosis as it relates to current psychotherapy practicesAn investigation that focuses on the relationship between the number of hours of television watching and gradepoint averageA case study analysis of the effectiveness of policies for educating all childrenThe effect of a preschool language program on the language skills of inner-city childrenGender differences in spatial and verbal abilitiesThis descriptive study provides information about the content of their online behaviors but tells you little about why they may do what they do. In this type of a research endeavor, you are not trying to understand the motivation for using what online tools are used nor are you trying to manipulate their use or content of the communication or any other outcome. This is nonexperimental in nature because no cause-and-effect relationships of any type are being hypothesized or investigated.Nonexperimental research methods that will be covered in this volume are descriptive, correlational, and qualitative. Descriptive and correlational methods will be covered in Chapter 9, and qualitative methods will be discussed in Chapter 10. The following is a brief overview of each.
Descriptive research describes the characteristics of an existing phenomenon. The every 10-year U.S. Census is an example of descriptive research as is any survey that assesses the current status of anything from the number of faucets in a house to the number of adults over 60 years of age who have grandchildren.Descriptive research focuses on events that occur in the present.What can be done with this information? First, it provides a broad picture of a phenomenon you might be interested in exploring. For example, if you are interested in learning more about the reading process in children, you might want to consult The Reading Report Card (at http://nces.ed.gov/nationsreportcard/reading/). This annual publication summarizes information about the reading achievement of children ages 9, 13, and 17 years. Or you might want to consult a publication of the Centers for Disease Control and Prevention, the Morbidity and Mortality Weekly Report (at http://www.cdc.gov/mmwr/), to determine the current incidence of measles cases in the Midwest, or the Bureau of Labor Statistics (at http://www.bls.gov/) to determine the current unemployment rate and the number of working single parents who have children under age 5 (about 60%). If you want to know it, there is a place to find it. Descriptive research demands this type of information.In another example, Eleanor Hanna, Hsiao-ye Yi, Mary Dufour, and Christine Whitmore (2001) examined the relationship of early smoking to alcohol use, depression, and drug use in adolescence. They used descriptive statistics and other statistical techniques to find that in comparison with those who never smoked, or those who simply experimented, early smokers were those most likely to use alcohol and other drugs as well as have school problems and early sexual experiences culminating in pregnancy.Descriptive research can stand on its own, but it can also serve as a basis for other types of research in that a group’s characteristics often need to be described before the meaningfulness of any differences can be addressed. And almost always descriptive data is collected but as the first step of many on the way to a more complex study. Want to describe an outcome? Learn about descriptive techniques.
Descriptive and historical research provide a picture of events that are currently happening or have occurred in the past. Researchers often want to go beyond mere description and begin discussing the relationship that certain events might have to one another. The most likely type of research to answer questions about the relationship among variables or events is called correlational research.What correlational research does, which neither descriptive nor historical research does, is to provide some indication as to how two or more things are related to one another or, in effect, what they share or have in common, or how well a specific outcome might be predicted by one or more pieces of information.Correlational research examines the relationship between variables.Correlational research uses a numerical index called the correlation coefficient (see Chapter 9 for a complete discussion) as a measure of the strength of this relationship. Most correlational studies report such an index when available.If you were interested in finding out the relationship between the number of hours that first-year students spend studying and their gradepoint averages, then you would be doing correlational research, because you are interested in the relationship between these two variables. If you were interested in finding out the best set of predictors of success in graduate school, you would be doing a type of correlational research that includes prediction.For example, in a study of culture, obesity stereotypes, self-esteem, and the “thin ideal,” Klaczynski, Goold, and Mudry (2004) examined the relationships among negative stereotypes of obesity, and other variables such as perceptions of the causes of obesity and of control over weight and self-esteem. They found a negative correlation between beliefs in control over one’s weight and self-esteem.One of the most important points about correlational research is that while it examines relationships between variables, it in no way implies that one causes changes in the other. In other words, correlation and prediction examine associations but not causal relationships, wherein a change in one factor directly influences a change in another.For example, it is a well-established fact that as the crime rate in a community increases, so does the level of ice cream consumption! What’s going on? Certainly, no rational person would conclude that the two are causally related such that if ice cream were banned, no more crimes would occur. Rather, another variable, temperature, better explains the increased ice cream consumption and the increased crime rate (both rise when it gets warm). It might seem ridiculous that people would identify causality just because events are related, but you do not have to read far in the daily newspaper to discover that politicians can reach just such unwise conclusions.
Qualitative research methods (see Chapter 10) are placed in this general category of nonexperimental methods because they do not directly test for cause and effect and, for the most part, follow an entirely different paradigm than the experimental model.Qualitative research studies phenomena within the social and cultural context in which they occur.The general purpose of qualitative research methods is to examine human behavior in the social, cultural, and political contexts in which they occur. This is done through a variety of tools, such as interviews, historical methods, case studies, and ethnography, and it usually results in qualitative (or nonnumerical) primary data. In other words, the qualitative researcher is more (but not only) interested in the contents of an interviewee’s speech than in the number of times (frequency) a particular comment is made.Qualitative research is relatively new to the social and behavioral sciences and, to a large extent, its increasing popularity is due to a degree of dissatisfaction with other available research methods. Some scientists feel that the traditional experimental model is too restrictive and narrow, preventing underlying and important factors and relationships from being revealed. What’s so valuable about this set of tools is that it allows you to answer a whole new set of questions in a whole new way.
You already know that correlational research can help to establish the presence of a relationship among variables, but it does not provide any reason to believe that variables are causally related to one another. How does one find out if characteristics, behaviors, or events are related in such a way that the relationship is a causal one? Two types of research can answer that question: true experimental research and quasi-experimental research.Experimental research examines the cause-and-effect relationship between variables.
True Experimental Research
In the true experimental research method, participants are assigned to groups based on some criterion, often called the treatment variable or treatment condition. For example, let us say that you are interested in comparing the effects of two different techniques for reducing obsessive-compulsive behavior in adults. The first technique includes behavioral therapy, and the second one does not. Once adults are assigned to groups and the programs are completed, you will want to look for any differences between the two groups with regard to the effects of the therapy on the frequency of obsessive-compulsive behaviors. Because the nature of the groups is determined by the researcher, the researcher has complete control over the factors to which the adults are exposed.True experimental research examines direct cause-and-effect relationships.This is the ideal model for establishing a cause-and-effect relationship because the researcher has clearly defined the possible cause (if indeed it results in some effect) and can keep very close tabs on what is happening. Most important, however, the researcher has complete control over the treatment.In a quasi-experimental study, the researcher does not have such a high degree of control because people have already been indirectly assigned to those groups (e.g., social class, type of abuse, gender, and type of injury) for which you are testing the effects.The distinction between experimental and other methods of research boils down to a matter of control. True experimental research designs (discussed in Chapter 11) isolate and control all the factors that could be responsible for any effects except the one of most interest.For example, Fleming, Klein, and Corter (1992) examined the effects of participation in a social support group on depression, maternal attitudes, and behavior in new mothers. As part of the experimental design, the researchers divided 142 mothers into three groups. Group 1 received the intervention, Group 2 received the no-intervention condition, and Group 3 received a special group-by-mail intervention. The key point here is the manipulation (the key word in experimental designs) of the condition for each of the three groups. This research is true experimental because the researchers determined the nature of the treatment and who is assigned to each group. As you will learn, in a quasi-experimental study, the researcher has no control over the origin of group membership (male or female, black or white, etc.). The primary difference between quasi-experimental and true experimental research is that in the former, subjects are preassigned to groups. It’s that simple.
In quasi-experimental research, participants are preassigned to groups based on some predetermined characteristic or quality. Differences in gender, race, age, grade in school, neighborhood of residence, type of job, and even experiences are examples. These group assignments have already taken place before the experiment begins, and the researcher has no control over who is assigned to which group.Quasi-experimental studies also focus on cause and effect, but they use preassigned groups.Let us say that you are interested in examining voting patterns as a function of neighborhood. You cannot change the neighborhood people live in, but you can use the quasi-experimental method to establish a causal link between residence and voting patterns. In other words, if you find that voting pattern and residence are related, then you can say with some degree of confidence (but not as much as with an experimental study) that there is a causal relationship between where one resides and how one votes.The most important use of the quasi-experimental method occurs where researchers cannot, in good conscience, assign people to groups and test the effects of group membership on some other outcome. For example, researchers who are interested in reducing the impact of child abuse cannot “create” groups of abusers, but rather have to look at already established groups of people who are abusive. That’s exactly what Mark Chaffin and his colleagues (2004) did when they assigned already (and that’s the key word) physically abusive parents to one of three intervention conditions. They found a reduction in abusive behavior by parents who were assigned to parent–child interaction therapy.Quasi-experimental research is also called post hoc, or after the fact, research because the actual research takes place after the assignment of groups (e.g., abusive versus nonabusive, employed versus unemployed, malnourished versus nonmalnourished, and male versus female). Because assignment has already taken place, the researcher has a high degree, but not the highest degree, of control over the cause of whatever effects are being examined. For the highest degree of control to occur, the true experimental model must be followed.Another phrase for quasi-experimental research is post-hoc, or after the fact.
We have briefly defined and discussed the different research methods that you will learn about later in Exploring Research in much greater detail. For now, answer this question. What determines the research method that a scientist should use to answer a question or test a hypothesis? Which research method described here best lends itself to questions you want answered?
What Research Method to Use When?
This is a beginning course and no one would expect you to be able to identify what type of research method was used in a particular study—at least not yet. You may have a very good idea if you understand what you just read about nonexperimental and experimental research methods, but it takes some experience to become really good at the identification process.So, here is a little jump start in the form of a “cheat” sheet (shown in Figure 1.2). This is not a substitute for learning how to distinguish nonexperimental from experimental research designs—it’s just a good way to get started and a bit of a help when you need it. Note that an alternative to any nonexperimental method is a qualitative approach (which is not shown in Figure 1.2).
Basic Research Versus Applied Research
Sometimes in the research world, distinctions must be made not only about the type of research but also about the most general category into which the implications or utility of the research might fall. This is where the distinction between basic and applied research comes in. But beware! This distinction is sometimes used as a convenient way to classify research activity rather than to shed light on the intent or purpose of the researcher and the importance of the study.The most basic distinction between the two types of research is that basic research (sometimes called pure research) is research that has no immediate application at the time it is completed, whereas applied research does. If this appears to be a somewhat ambiguous distinction, it is, because almost all basic research eventually results in some worthwhile application over the long term. In fact, the once easy distinction between the two is slowly disappearing.Both basic and applied research are critical parts of studying and understanding a wide range of phenomena.
For example, for every dollar spent on the basic research that supported the lunar missions during the 1960s and 1970s, $6 were returned in economic impact. Data from basic research that hypothesizes a relationship between Alzheimer’s disease in older people and Down’s syndrome (a genetic disorder) in younger people could eventually prove to be the critical finding that leads to a cure for both conditions. Another example: Who cares if some children have a more difficult time than others do in distinguishing between two very similar stimuli? You do, if you want to teach these children how to read. Many different reading programs have grown directly from such basic research efforts.Never judge the quality of either the finished product or the worth of supporting a research project by branding it as basic or applied research. Rather, look closely at its content and judge it on its merit. This approach obviously has been used, because more and more reports about basic research (at one time beyond the interests of everyday practitioners) appear in such practitioner-oriented professional journals as Phi Delta Kappan and the APA Monitor, as well as the Sunday New York Times Magazine, Newsweek, Science News, and American Scientist. And the results of applied research are those that policy makers look to when formulating position papers.
Why are both basic and applied research essential to the scientific community as well as to the public community that it serves? What do you think an educated or informed citizen should know about how the research process works? What five questions might he or she be able to answer?
Great! You have finished the first chapter of Exploring Research, and hopefully you now have a good idea about what research is (and isn’t), what the purpose of research is, and some of the different ways in which research can be carried out. With this new information under your belt, let’s turn to the next chapter, which focuses on some “researchese,” or the language used by researchers, and how these new terms fit together with what you have learned here.
1.The process of research never stands independently from the content of the research. As a student new to the field of research, and perhaps even to your own discipline (such as education, psychology, sociology, or nursing), answer the following questions:
- (a)What areas within your discipline especially interest you?
- (b)Who are some of the outstanding researchers in your field, and what is the focus of their work?
- (c)Of the different types of research described and discussed in this chapter, which one do you think best fits the type of research that is done in your discipline?
2.At this point in your studies, what do you find most intimidating about the research process? What is one thing you could do to make this part of the research process a little bit easier or more comfortable? In which part of conducting research are you most confident?3.How do the terms “hypothesis” and “theory” differ in meaning?4.Visit your college or university library and locate an article from a professional journal that describes a research study. Access it online, or as a hard copy. From the description of how scientific inquiry takes place (which you read about in this chapter), answer the following:
- (a)What is the primary question posed by the study?
- (b)What important factors are identified?
- (c)Is there a hypothesis stated? If so, what is it?
- (d)Describe how the information was collected.
- (e)How do the results of the study affect the original hypothesis?
5.Interview an active researcher on your campus and ask about this person’s research activities, including:
- (a)The focus of this person’s research interests.
- (b)Why this individual is interested in this area.
- (c)What the most exciting part of the research is.
- (d)What the least exciting part of the research is.
- (e)What impact the results of the research might have on this individual’s particular discipline.
- (f)What studies this individual would like to see as follow-up studies to the research.
6.Select a discipline within the social and behavioral sciences, such as child development, social psychology, higher education, or health psychology. For the discipline you select, find a representative study that is quasi-experimental or experimental in nature. Write a one-paragraph description of the study. Do the same for a historical study.7.This chapter contains several examples of preassigned groups used in quasi-experimental research (e.g., groups based on preassignment such as gender, race, grade in school, etc.). Name three more examples of preassigned groups appropriate for quasi-experimental research.8.Research questions come from imagination and can be enriched by science, art, music, and literature. Identify a book you have read or a television show or movie you have watched. What kind of research question can you pull from this work? Here are some examples to get you started:“Pride and Prejudice” (Jane Austen): In what ways do perceptions of social status relate to choices in a relationship partner?“Clueless” (Amy Heckerling): How does an intervention involving vocabulary lessons, a new wardrobe, and instructions on which social groups to befriend affect ratings of popularity from fellow high school students?9.Find a normal part of your daily routine about which to ask an “I wonder if . . . ” question. For example, “I wonder if the amount of text messaging in the hour before bedtime affects the amount of time needed to fall asleep in adolescents.”10.In a fictitious correlational study, the results showed that age was related to strength, that is, as children get older, their strength increases. What is the problem with the statements that increased strength is caused by increasing age, or that the stronger you get the older you get?11.Write down your definition of science. How would your definition of science differ from a student’s in a similar class 25 years ago? How would your definition differ from that put forth by a physical (e.g., physics, chemistry) scientist, if it differs at all?12.When trying to decide which scientific method to use when exploring a question, what is the best rule of thumb to go by?13.Look for examples of editorials or research articles that present correlational evidence. Do the authors infer a cause-and-effect relationship in the correlation? Why might it be difficult for even seasoned researchers to avoid making this mistake?14.Research often replicates findings made by others. What is the value in this process?15.We live in a very complex world just filled with economic and social challenges. How can the research process help us solve or better understand some of those problems and issues?16.Identify five attributes that characterize high-quality research.17.A researcher who hypothesized that 6-year-old children of nonworking mothers have more advanced reading skills than those of 6-year-old children of working mothers found insignificant results. Based on this information and what you have learned about the field of research, answer the following questions:
- (a)What is a new research question the researcher could ask?
- (b)What is one step in between examining the results and asking the new research question that might point the researcher in the right direction?
18.Two characteristics of high-quality research are generalizability and the ability to contribute to the betterment of society. In other words, results from high-quality research, particularly applied research, can provide a meaningful answer to the question, “So What?” Read a research article and describe in one or two sentences how the research addresses the “So What?” question.19.Explain the difference between historical, correlational, and quasi-experimental research.20.Here’s the question . . .What is the difference between achievement scores for a group of children born inPeoria and a group born in Croatia?Use Figure 1.2 to determine the method you should use.
Online. . .
Because someday you’ll be a professional, there’s no time like the present to get information about some professional societies and join as a student—it will never be cheaper. Here are some of the largest organizations and their Internet addresses:
- • American Anthropology Association at http://www.aaanet.org/
- • American Educational Research Association at http://www.aera.net/
- • American Medical Association at http://www.ama-assn.org/
- • American Psychological Association at http://www.apa.org/
- • American Public Health Association at http://www.apha.org/
- • National Association for the Education of Young Children at http://www.naeyc.org/
- • American Nurses Association at www.nursingworld.org/
- • American Association for the Advancement of Science at www.aaas.org/
- • American Statistical Association at http://www.amstat.org
- • American Psychiatric Association at http://www.psych.org
- • American Pharmacists Association at http://www.pharmacist.com
- • Council for Exceptional Children at http://www.cec.sped.org