|
Michael Biderman's Web Site
|
2004 Graduate School Thesis Workshop videos.Research IdeasFollowing are suggestions based on the research of my colleagues and me upon which you might build your own research project. They're listed in approximate chronological order of the research that lead them. If you have questions about the project or want to "shoot the bull" about one of them, feel free to email me at Michael-Biderman@utc.edu or call me at 423-425-4268. Outgoing message gives more phone numbers. Some papers upon which the research suggestions are based can be found at recent papers that can be downloaded .
Quionna Caldwell's Thesis. Who appreciates organizational diversity efforts? Quionna found that the relationship of Affective Commitment and Intent to Leave were related to organizations' diversity efforts. She used a sample of African-American female managers. She found that those managers who perceived their organizations as engaging in stronger diversity efforts were more committed and less likely to leave than those who perceived their organizations as engaging in fewer diversity efforts. More importantly, she found that the relationships were significantly stronger among those managers who expressed high value for diversity efforts. Quionna used a single ethinic group, single gender sample. This study should be replicated and extended using African-Americans and Whites, males and females. I'm interested in the extend to which Value for Diversity Efforts moderates the relationships among all ethnic and gender groups. I'm also interested in determining if Value for Diversity Efforts might be a mediator of Gender differences. See the Caldwell, et. al. SIOP paper in recent papers that can be downloaded for more details on her study.
Predicting Turnover. How can we select employees who won't leave the organization? As part of a study of turnover at McKee Baking, we gathered data on several hundred manufacturing employees. We looked at the relationship of several variables available at the time the employees applied for work to how long the employees stayed with the company. We found some variables that were significantly related to turnover. One of the most interesting, yet vexing, was score on a manual dexterity test. The applicants who performed the test most quickly were also most likely to leave the organization. This relationship held up after controlling for a variety of other variables, including age and gender and type of job. We also found that applicants who had higher appearance ratings during the pre-employment interview were less likely to leave. This study was conducted a couple of years ago, when the economy was at nearly full employment. I'm interested in knowing whether the relationships of turnover to the variables we studied are the same now that unemployment is high. See the Biderman, et. al. SIOP paper in recent papers that can be downloaded for more details.
Terri Rieth's dissertation. What makes an effective organization? Dr. Terri Rieth, a 1996 graduate of the program, conducted a study investigating factors related to organizational effectiveness. She used physical therapy clinics as the unit of analysis, sending questionnaires to over 100 clinics in Tennessee, Georgia, and Alabama. She obtained usable data from over 500 employees of 87 clinics. She found that those clinics in which the boundaries between different levels of authority were more permeable were more effective than those in which the boundaries were more rigidly enforced. Because it involved relationships between individual employees and between the clinics in which they work, the study required the use of special analytic techniques called hierarchical linear modeling. This is the first study to investigate the relationship of organization effectiveness to aspects of the boundaries between levels of authority. It would be interesting to see if the results obtained by Dr. Rieth hold up in different types of organizations. See the Rieth and Biderman. SIOP paper in recent papers that can be downloaded for more details.
Nhung Nguyen's dissertation. Is there a general faking ability? In a collaboration with a 1998 graduate of the program, Dr. Nhung Nguyen, a structural equation model of faking ability was developed. The model was applied to data of Nhung's dissertation. Those data were from an "Instructed faking" design, in which participants were given tests twice, once under instructions to respond honestly and then again under instructions to fake good. A questionnaire covering the Big 5 personality dimensions and a situational judgment test for customer service representatives were administered. The application of the model provided evidence that there is a generalized ability to fake such tests. A Faking Ability factor loaded significantly on all tests, including the situational judgment test. We also found that faking ability was positively related to cognitive ability and that it was not related to any of the Big 5 personality dimensions. A SIOP paper describing the application of the model is on the recent papers that can be downloaded page. We're interested in the extent to which faking ability is related to scores on other types of situational judgment tests. Jody Damron's thesis. Can people fake toward specific jobs? Jody administered the Big 5 questionnaire to four groups of participants. Each group was instructed to respond to the Big 5 questionnaire honestly, then to respond again as if they were applying for a specific job. The four groups differed in terms of the job to which they were applying. One group was given the job description of a manager. The 2nd was given the job description of an accountant. The 3rd was given the description of a Sales person. The 4th group was simply told to "fake good". He found very few difference in the profile of Big 5 scores across the four conditions. This suggests that persons aren't very good at faking toward specific jobs. Jody's manipulation was intended to mimic the kind of information applicants would have in the real world - descriptions of the jobs to which they were applying. A natural follow-up to this study would be to strengthen the manipulation - to give participants stronger indications of the kinds of profiles appropriate for each of the jobs. If this follow-up also found no difference in profiles across jobs applied for, that result would suggest that persons hiring for different jobs need only worry about generalized faking, and need not be concerned about faking specific to a job. On the other hand, the finding of job-specific faking would require that persons hiring for particular jobs look for specific kinds of faking. Lyndsay Wrensen's independent study. Who is good at faking? Lyndsay administered the Big 5 questionnaire along with several other personality and cognitive tests. She used the "instructed faking" design employed in all the studies considered here so far. She found that persons who faked best were those with high cognitive ability, high emotional intelligence, and high integrity who gave fewest socially desirable responses. She found no relationship of faking ability to the Big 5 dimensions after controlling for cognitive ability. There are several follow-ups to this study. Many of them would involve looking at other measures of social desirability, emotional intelligence, and integrity, in order to pinpoint more precisely the personality profile of good fakers. Others would involve using a design that involved incentives to fake rather than instructions to fake. Mike Clark's thesis. Is propensity to fake different from ability to fake? Mike is going to attempt to measure the tendency or propensity to fake separately from the ability to fake. All of the studies above have used the "instructed faking" design. In this design, participants are instructed to fake good. Thus, persons who would normally not fake (low tendency) and those who would normally be quite likely to fake (high tendency) are all told to fake. In this design it is assumed that all persons had the same tendency to fake, since they were told that it was expected of them. But, of course, this is not the way the world is. There are some people who will fake to get ahead (high tendency) while others respond honestly, in spite of advantages faking might give them (low tendency). Mike will employ conditions and measures in which hopefully will allow him to measure faking tendency separately from faking ability. A natural follow-up to Mike's thesis would be to create an artificial job situation that students could apply for and then examine the effect of differences in faking tendency and faking ability on validity of predictors of job performance. Another follow-up is to develop alternative measures of faking tendency and faking ability, measures that could be used in an employment setting and that might then be used to adjust applicant scores. Jennifer Sebren's independent study. Why do conscientious people get better grades? Jennifer administered the Big 5 personality inventory to a group of PSY 101 students. She also had them fill out a questionnaire about their study habits. Our hypothesis was that students high in C are driven to study more than those low in C. Longer times spent studying lead to higher grades. We got the 2nd part of the mediation hypothesis: Students who self reported studying more got better grades. Interestingly, the 1st part of the mediation hypothesis - the link between Conscientiousness and study time was not significant when we used raw Conscientiousness scale scores as the measure of C. But when we applied a structural equation model and defined Conscientiousness using factor scores derived from the structural equation model, everything fell into place. Conscientiousness was related to study time. Study time was related to performance. And when the C -> Study time -> Performance link was controlled for, the relationship of C to performance was not significant. This suggested that the main, perhaps the sole reason that Conscientiousness predicts performance in academic settings is through its relationship to study time. Mike Clark's thesis again. How do people fake? In Mike's thesis, we found that while participants made generallly higher scores in the faking conditions, several participants seemed to engage in what might be characterized as targeting behavior - restricting most of their responses to one of the response alternatives. The nature of faking behavior has implications for the models of faking that are appropriate, so a study examining factors that might elicit this targeting behavior or prevent it is in order. Biderman & Nguyen, 2006. Measuring faking in an applicant sample. In a paper presented in 2006, we speculated that it might be possible to measure both scores on the Big 5 dimensions and a faking score from a single administration of the Big 5. We need an applicant sample of 200 or more persons to test this hypothesis. Somebody do this, please!!!
|