The purpose of this study was to determine whether there was a correlation between guilt and sham... more The purpose of this study was to determine whether there was a correlation between guilt and shame proneness and frequency with self reported impulsivity. This study utilized Barratt Impulsiveness Scale, Harder's Personal Feelings Questionnaire, and Eyre's (2004) Shame and Guilt Inventory to measure the responses of participants. Results indicated that participants did not discriminate well between the guilt and shame subscales and that the two measures of guilt and shame were only moderately correlated with each other. However, the PFQ-2 shame frequency was moderately correlated with overall impulsiveness and each of the impulsiveness subscales. PFQ-2 guilt frequency was moderately related to overall impulsivity and non-planning impulsiveness. The SAGI measure of guilt and shame proneness was unrelated to any of the impulsivity scales. Future research should investigate why impulsiveness is related to shame frequency more than guilt frequency and why proneness to guilt and shame does not predict an individual's trait impulsiveness level.
College students completed the Khavri Alcohol Test to assess their Annual Absolute Alcohol Intake... more College students completed the Khavri Alcohol Test to assess their Annual Absolute Alcohol Intake (AAAI), a computerized measure of delayed discounting (impulsivity), and had their height/weight measured to calculate BMI. The correlation between these measures and whether discounting rates varied for different categories of participants (e.g., heavy drinking/obese) is examined.
Abstract: The purpose of this research was to develop a new scenario-based measure of shame-and g... more Abstract: The purpose of this research was to develop a new scenario-based measure of shame-and guilt-proneness entitled the Shame and Guilt Inventory (SAGI). Studies 1 and 2 were designed to narrow down an initial pool of possible scenarios generated by ...
Now that we have taken the first step to classify each open-ended response on the ASC, in future ... more Now that we have taken the first step to classify each open-ended response on the ASC, in future endeavors we plan to determine whether responses to the ASC will predict responses on other trait emotion measures, measures of psychopathology, and behavioral measures. Limitations of the current study include that although the authors made every effort to be unbiased in their codings of the open-ended responses, there is the possibility that some systematic bias creeped in. In conclusion, the ASC has potential to revive the interest in semi-projective measures of emotion, although future research will be needed to see if these classifications can be easily used and understood by other researchers. Results To analyze the data, first we selected the sentence-stem classification categories in which more than 10% of the students responses were coded. Second, we selected the emotion words that were chosen by more than 9% of the students. Next, chi-square analyses were used to determine if there was a relationship between the type of response given to the sentence stem and the emotion selected for the "this makes me feel" prompt. We found that there was not a statistically significant relationship for eight of the 39 sentence stems. However, for the remaining 31 stems, the chi-square results were statistically significant. Eight of these stems, their response classification, and the emotion chosen are depicted below. Method Undergraduates from two regional public universities (N = 536) completed a 39question version of the Affective Sentence Completion test (Ferguson, Sorenson, & Eyre, 1996) in addition to several other measures of guilt and shame that are not analyzed in the current presentation. There were 202 Males (39%) and 323 Females (61%); of these, 155 were African American or Black (29%), 340 were Caucasian or White (64%), and 28 were other ethnicities or races (7%). Most were Freshmen (55%) or Sophomores (23%). The first two authors, in conjunction with their faculty advisor, then created categories to code each open-ended question and then established interrater reliabilities on 6% to 16% of the questions (Kappas ranged from .564 to 1.00). The third author, blind to the goals of the study, then did a secondary reliability check on 4% to 8% of the questions (Kappas ranged from on .423 to .941).
The current study examined the relationship between delayed discounting of hypothetical money and... more The current study examined the relationship between delayed discounting of hypothetical money and effort discounting of hypothetical extra credit points. Both money and extra credit showed evidence of magnitude effects, such that larger amounts were discounted less than smaller amounts. With regard to stability of answers, there were high correlations between all of extra credit discounting questions and their duplicates, but not for the monetary discounting measures. The discussion speculates as to why money showed different results than extra credit.
• The goal of the current study was to examine the relationship between study distractors, procra... more • The goal of the current study was to examine the relationship between study distractors, procrastination, and impulsivity, and relate these to course performance in psychology courses and overall GPA. Results indicated that impulsivity, procrastination, and study distractors were not generally related to course performance or GPA. However, impulsivity and procrastination were significantly related to a number of common study distractors that students might encounter (e.g., other people talking, TV, internet, amount of coursework). Future research should examine why impulsivity, study distractors, procrastination were not good predictors of course performance or GPA.
Research on temporal discounting has primarily used hypothetical monetary rewards. For the colleg... more Research on temporal discounting has primarily used hypothetical monetary rewards. For the college student population, course points serve as a powerful generalized reinforcer, much like money does for the general population. To date, few studies have examined the degree to which students discount points within a course. Silva and Gross (2004, Study 2) found that higher performing students were less likely to discount extra credit points than middle or lower performing students. However, this study focused on the effort to earn the reinforcer, not on the delay of the reinforcer. Williams and colleagues (2007) developed a computerized procedure to examine the rate at which students discounted extra credit points. They found that the majority of students did not discount course points and those who did showed rather shallow discounting curves. They posited that these results may have been due to a magnitude effect (10 course points for a maximum of 55 minutes) or too short of delay periods (ranging from 1 minute to 10 minutes). Further, participants might have been more willing to engage in self-control behaviors on the longer time delays due to prior commitment on the shorter time delays. The current study was designed to address some of the aforementioned limitations in Williams et al. (2007) by lengthening the delays, counterbalancing the presentation order, and looking for evidence of a magnitude effect.
Although impulsivity has traditionally been studied using standard personality inventories such a... more Although impulsivity has traditionally been studied using standard personality inventories such as the Barratt Impulsiveness Scale (BIS; Patton, Stanford, & Barratt, 1995), a recent trend in behavior analysis and behavioral economics has been to look to temporal discounting as a measure of impulsiveness. Temporal discounting refers to people's tendency to prefer smaller-sooner rewards to larger-later rewards. That is, they devalue long-term outcomes in favor of short-term gains and losses. That is, in the economic domain, people often prefer to receive some smaller amount of money today (e.g., $50) rather than waiting 1 month to receive $100. The smaller-sooner choice is seen as impulsive because of the long-term detrimental outcome of the choice. This procedure has been used extensively to assess the relationship between impulsiveness and addictive behaviors such as smoking, drug use, gambling, and obesity (e.g., Yi, Mitchell, & Bickel, 2010). However, to date, few studies how examined temporal discounting relates to traditional personality measures. For example, research has found that people who are more extraverted tend to show steeper discounting rates compared to people who are more introverted, but sensation seeking does not appear to be related to discounting (e.g., Ostaszewski, 1997). Another study examined economic locus of control (the belief that one's economic outcomes are due to chance/external factors or one's own actions) in relationship to temporal discounting (Plunkett & Buehner, 2007). With regard to the relationship between the BIS and temporal discounting, some studies have found a relationship, while others have not (de Wit, Flory, Acheson, McCloskey, & Manuck, 2007). The purpose of the present study is to examine several personality variables that should be related to temporal discounting if it is indeed a measure of impulsivity. Specifically, 172 general psychology students completed an online survey that contained Rotter's Internal-External Scale (Rotter, 1966), the Short Form of the Need for Cognition Scale (Cacioppo, Petty, & Kao, 1984), Steel, Brothen, and Wambach's (2001) procrastination measure, and the BIS-11 (Patton et al., 1995). Approximately one week after completing the survey, 132 students came to the laboratory and completed a computerized measure of temporal discounting which asked them to choose between a maximum future amount (e.g., $100 in one month) and an adjusting immediate amount (e.g., $50 now). This amount was titrated until participants reached their indifference point (the smallest amount they would prefer to receive now, rather than wait for the larger-later reward). This discounting procedure employed three magnitudes of reinforcers
The study of temporal discounting of rewards-the tendency to prefer a smaller, sooner reward over... more The study of temporal discounting of rewards-the tendency to prefer a smaller, sooner reward over a larger, later reward-has been one of the topics examined in the greatest depth in the behavioral economics literature, with probability discounting of rewards-the tendency to prefer rewards that have a higher probability of occurring-following close behind. However, both of these constructs refer to rewards that are received in cases where the participant has put forth no effort. Except for cases such as investment or gambling, little comes to us without effort. One area that has received less empirical study is the domain of effort discounting (e.g, Silva & Gross, 2004). One reason for this may be because effort and time are confounded. This is especially true in education, where students must actively engage in reading and comprehending course material in order to receive a high score. But, even though time and effort are inexorably intertwined in the education field, people can be asked to make a decision about where their indifference point lies with regard to how much effort/time they would be willing to expend in order to receive differing exam grades. The purpose of the present study was to examine whether students would discount exam grades in a hyperbolic-like manner, showing the same type of discounting curve found in temporal and probability discounting, and to correlate exam discounting with the discounting of monetary rewards. Method Participants For this presentation, we are using two data sets to replicate the findings. The first data set was collected in Spring 2012 (hereafter referred to as "Spring 2012"). In this set there were 209 participants from introductory and advanced psychology courses who ranged in age from 18 to 54 years old, with a mean age of 22.6 years. Thirty-eight percent were male and 62% were female. The majority of students were second semester freshmen (49%), but they ranged from first semester freshmen (9%) to 5 th year senior or beyond (2%). The second data set was collected during several semesters in 2011 and 2012 (hereafter referred to as "2011-2012"). This sample consisted of 318 participants from introductory and advanced psychology courses who ranged in age from 18 to 56 years old, with a mean age of 22.7. Males made up 35% of the sample and females made up 62% (2% were missing demographic information). Again the majority of students were second semester freshmen (41%). Materials and Procedure Participants completed an online survey in exchange for course credit. Several personality scales were included along with some individual difference measures, but they are not included in the current analysis. Measures that are included in the analysis are described below. Exam Discounting Measure Participants were asked a series of questions which were identical in wording, with the exception that the maximum number of hours required to study to receive 100% was altered to be 1,
Kelley’s (1967) covariation theory as operationalized by Orvis, Cunningham, and Kelley’s (1975) i... more Kelley’s (1967) covariation theory as operationalized by Orvis, Cunningham, and Kelley’s (1975) is considered one of the prevailing models of causal attribution. With one notable exception (McArthur, 1972), no one has examined the attribution of emotion within this framework. The present study extended the findings of McArthur by having participants make causal attributions for four emotions (anger, happiness, depression, and anxiety) embedded within an interpersonal context. The sample consisted of 155 male and female native English-speaking Canadian undergraduates. Results supported Orvis et al.’s (1975) predictions for the person and stimulus attributions. There was less support for the circumstance attributions, and the least amount of support was found for the person x stimulus attribution. Discussion focused on design and analysis limitations, applications of the current study to other realms in emotion, and the need for a more extensive examination of other emotions.
Research on temporal discounting has primarily used hypothetical monetary rewards. For the colleg... more Research on temporal discounting has primarily used hypothetical monetary rewards. For the college student population, course points serve as a powerful generalized reinforcer, much like money does for the general population. To date, few studies have examined the degree to which students discount points within a course. Silva and Gross (2004, Study 2) found that higher performing students were less likely to discount extra credit points than middle or lower performing students. However, this study focused on the effort to earn the reinforcer, not on the delay of the reinforcer. Williams and colleagues (2007) developed a computerized procedure to examine the rate at which students discounted extra credit points. They found that the majority of students did not discount course points and those who did showed rather shallow discounting curves. They posited that these results may have been due to a magnitude effect (10 course points for a maximum of 55 minutes) or too short of delay periods (ranging from 1 minute to 10 minutes). Further, participants might have been more willing to engage in self-control behaviors on the longer time delays due to prior commitment on the shorter time delays. The current study was designed to address some of the aforementioned limitations in Williams et al. (2007) by lengthening the delays, counterbalancing the presentation order, and looking for evidence of a magnitude effect.
Previous studies have shown that disgust sensitivity (DS), right-wing authoritarianism (RWA), rel... more Previous studies have shown that disgust sensitivity (DS), right-wing authoritarianism (RWA), religious fundamentalism (RF), and belief in a just world (BJW) are associated with social conservatism and religious questioning (QUEST) is associated with social liberalism. The current study was designed determine whether these personality traits would predict who the student planned to vote for in the 2016 U.S. Presidential primary. Consistent with previous research, DS, RWA, RF, and BJW were positively correlated with social conservatism and QUEST was negatively correlated. Students who were highest on RWA, RF, and lowest on QUEST were most likely to vote for one of the Republican candidates. In contrast, students who were at the opposite end of the scales were most likely to vote for Bernie Sanders, with Hillary Clinton supporters in the middle. There were no differences in DS and few differences in BJW with regard to preferred candidate.
The purpose of this study was to determine whether there was a correlation between guilt and sham... more The purpose of this study was to determine whether there was a correlation between guilt and shame proneness and frequency with self reported impulsivity. This study utilized Barratt Impulsiveness Scale, Harder's Personal Feelings Questionnaire, and Eyre's (2004) Shame and Guilt Inventory to measure the responses of participants. Results indicated that participants did not discriminate well between the guilt and shame subscales and that the two measures of guilt and shame were only moderately correlated with each other. However, the PFQ-2 shame frequency was moderately correlated with overall impulsiveness and each of the impulsiveness subscales. PFQ-2 guilt frequency was moderately related to overall impulsivity and non-planning impulsiveness. The SAGI measure of guilt and shame proneness was unrelated to any of the impulsivity scales. Future research should investigate why impulsiveness is related to shame frequency more than guilt frequency and why proneness to guilt and shame does not predict an individual's trait impulsiveness level.
College students completed the Khavri Alcohol Test to assess their Annual Absolute Alcohol Intake... more College students completed the Khavri Alcohol Test to assess their Annual Absolute Alcohol Intake (AAAI), a computerized measure of delayed discounting (impulsivity), and had their height/weight measured to calculate BMI. The correlation between these measures and whether discounting rates varied for different categories of participants (e.g., heavy drinking/obese) is examined.
Abstract: The purpose of this research was to develop a new scenario-based measure of shame-and g... more Abstract: The purpose of this research was to develop a new scenario-based measure of shame-and guilt-proneness entitled the Shame and Guilt Inventory (SAGI). Studies 1 and 2 were designed to narrow down an initial pool of possible scenarios generated by ...
Now that we have taken the first step to classify each open-ended response on the ASC, in future ... more Now that we have taken the first step to classify each open-ended response on the ASC, in future endeavors we plan to determine whether responses to the ASC will predict responses on other trait emotion measures, measures of psychopathology, and behavioral measures. Limitations of the current study include that although the authors made every effort to be unbiased in their codings of the open-ended responses, there is the possibility that some systematic bias creeped in. In conclusion, the ASC has potential to revive the interest in semi-projective measures of emotion, although future research will be needed to see if these classifications can be easily used and understood by other researchers. Results To analyze the data, first we selected the sentence-stem classification categories in which more than 10% of the students responses were coded. Second, we selected the emotion words that were chosen by more than 9% of the students. Next, chi-square analyses were used to determine if there was a relationship between the type of response given to the sentence stem and the emotion selected for the "this makes me feel" prompt. We found that there was not a statistically significant relationship for eight of the 39 sentence stems. However, for the remaining 31 stems, the chi-square results were statistically significant. Eight of these stems, their response classification, and the emotion chosen are depicted below. Method Undergraduates from two regional public universities (N = 536) completed a 39question version of the Affective Sentence Completion test (Ferguson, Sorenson, & Eyre, 1996) in addition to several other measures of guilt and shame that are not analyzed in the current presentation. There were 202 Males (39%) and 323 Females (61%); of these, 155 were African American or Black (29%), 340 were Caucasian or White (64%), and 28 were other ethnicities or races (7%). Most were Freshmen (55%) or Sophomores (23%). The first two authors, in conjunction with their faculty advisor, then created categories to code each open-ended question and then established interrater reliabilities on 6% to 16% of the questions (Kappas ranged from .564 to 1.00). The third author, blind to the goals of the study, then did a secondary reliability check on 4% to 8% of the questions (Kappas ranged from on .423 to .941).
The current study examined the relationship between delayed discounting of hypothetical money and... more The current study examined the relationship between delayed discounting of hypothetical money and effort discounting of hypothetical extra credit points. Both money and extra credit showed evidence of magnitude effects, such that larger amounts were discounted less than smaller amounts. With regard to stability of answers, there were high correlations between all of extra credit discounting questions and their duplicates, but not for the monetary discounting measures. The discussion speculates as to why money showed different results than extra credit.
• The goal of the current study was to examine the relationship between study distractors, procra... more • The goal of the current study was to examine the relationship between study distractors, procrastination, and impulsivity, and relate these to course performance in psychology courses and overall GPA. Results indicated that impulsivity, procrastination, and study distractors were not generally related to course performance or GPA. However, impulsivity and procrastination were significantly related to a number of common study distractors that students might encounter (e.g., other people talking, TV, internet, amount of coursework). Future research should examine why impulsivity, study distractors, procrastination were not good predictors of course performance or GPA.
Research on temporal discounting has primarily used hypothetical monetary rewards. For the colleg... more Research on temporal discounting has primarily used hypothetical monetary rewards. For the college student population, course points serve as a powerful generalized reinforcer, much like money does for the general population. To date, few studies have examined the degree to which students discount points within a course. Silva and Gross (2004, Study 2) found that higher performing students were less likely to discount extra credit points than middle or lower performing students. However, this study focused on the effort to earn the reinforcer, not on the delay of the reinforcer. Williams and colleagues (2007) developed a computerized procedure to examine the rate at which students discounted extra credit points. They found that the majority of students did not discount course points and those who did showed rather shallow discounting curves. They posited that these results may have been due to a magnitude effect (10 course points for a maximum of 55 minutes) or too short of delay periods (ranging from 1 minute to 10 minutes). Further, participants might have been more willing to engage in self-control behaviors on the longer time delays due to prior commitment on the shorter time delays. The current study was designed to address some of the aforementioned limitations in Williams et al. (2007) by lengthening the delays, counterbalancing the presentation order, and looking for evidence of a magnitude effect.
Although impulsivity has traditionally been studied using standard personality inventories such a... more Although impulsivity has traditionally been studied using standard personality inventories such as the Barratt Impulsiveness Scale (BIS; Patton, Stanford, & Barratt, 1995), a recent trend in behavior analysis and behavioral economics has been to look to temporal discounting as a measure of impulsiveness. Temporal discounting refers to people's tendency to prefer smaller-sooner rewards to larger-later rewards. That is, they devalue long-term outcomes in favor of short-term gains and losses. That is, in the economic domain, people often prefer to receive some smaller amount of money today (e.g., $50) rather than waiting 1 month to receive $100. The smaller-sooner choice is seen as impulsive because of the long-term detrimental outcome of the choice. This procedure has been used extensively to assess the relationship between impulsiveness and addictive behaviors such as smoking, drug use, gambling, and obesity (e.g., Yi, Mitchell, & Bickel, 2010). However, to date, few studies how examined temporal discounting relates to traditional personality measures. For example, research has found that people who are more extraverted tend to show steeper discounting rates compared to people who are more introverted, but sensation seeking does not appear to be related to discounting (e.g., Ostaszewski, 1997). Another study examined economic locus of control (the belief that one's economic outcomes are due to chance/external factors or one's own actions) in relationship to temporal discounting (Plunkett & Buehner, 2007). With regard to the relationship between the BIS and temporal discounting, some studies have found a relationship, while others have not (de Wit, Flory, Acheson, McCloskey, & Manuck, 2007). The purpose of the present study is to examine several personality variables that should be related to temporal discounting if it is indeed a measure of impulsivity. Specifically, 172 general psychology students completed an online survey that contained Rotter's Internal-External Scale (Rotter, 1966), the Short Form of the Need for Cognition Scale (Cacioppo, Petty, & Kao, 1984), Steel, Brothen, and Wambach's (2001) procrastination measure, and the BIS-11 (Patton et al., 1995). Approximately one week after completing the survey, 132 students came to the laboratory and completed a computerized measure of temporal discounting which asked them to choose between a maximum future amount (e.g., $100 in one month) and an adjusting immediate amount (e.g., $50 now). This amount was titrated until participants reached their indifference point (the smallest amount they would prefer to receive now, rather than wait for the larger-later reward). This discounting procedure employed three magnitudes of reinforcers
The study of temporal discounting of rewards-the tendency to prefer a smaller, sooner reward over... more The study of temporal discounting of rewards-the tendency to prefer a smaller, sooner reward over a larger, later reward-has been one of the topics examined in the greatest depth in the behavioral economics literature, with probability discounting of rewards-the tendency to prefer rewards that have a higher probability of occurring-following close behind. However, both of these constructs refer to rewards that are received in cases where the participant has put forth no effort. Except for cases such as investment or gambling, little comes to us without effort. One area that has received less empirical study is the domain of effort discounting (e.g, Silva & Gross, 2004). One reason for this may be because effort and time are confounded. This is especially true in education, where students must actively engage in reading and comprehending course material in order to receive a high score. But, even though time and effort are inexorably intertwined in the education field, people can be asked to make a decision about where their indifference point lies with regard to how much effort/time they would be willing to expend in order to receive differing exam grades. The purpose of the present study was to examine whether students would discount exam grades in a hyperbolic-like manner, showing the same type of discounting curve found in temporal and probability discounting, and to correlate exam discounting with the discounting of monetary rewards. Method Participants For this presentation, we are using two data sets to replicate the findings. The first data set was collected in Spring 2012 (hereafter referred to as "Spring 2012"). In this set there were 209 participants from introductory and advanced psychology courses who ranged in age from 18 to 54 years old, with a mean age of 22.6 years. Thirty-eight percent were male and 62% were female. The majority of students were second semester freshmen (49%), but they ranged from first semester freshmen (9%) to 5 th year senior or beyond (2%). The second data set was collected during several semesters in 2011 and 2012 (hereafter referred to as "2011-2012"). This sample consisted of 318 participants from introductory and advanced psychology courses who ranged in age from 18 to 56 years old, with a mean age of 22.7. Males made up 35% of the sample and females made up 62% (2% were missing demographic information). Again the majority of students were second semester freshmen (41%). Materials and Procedure Participants completed an online survey in exchange for course credit. Several personality scales were included along with some individual difference measures, but they are not included in the current analysis. Measures that are included in the analysis are described below. Exam Discounting Measure Participants were asked a series of questions which were identical in wording, with the exception that the maximum number of hours required to study to receive 100% was altered to be 1,
Kelley’s (1967) covariation theory as operationalized by Orvis, Cunningham, and Kelley’s (1975) i... more Kelley’s (1967) covariation theory as operationalized by Orvis, Cunningham, and Kelley’s (1975) is considered one of the prevailing models of causal attribution. With one notable exception (McArthur, 1972), no one has examined the attribution of emotion within this framework. The present study extended the findings of McArthur by having participants make causal attributions for four emotions (anger, happiness, depression, and anxiety) embedded within an interpersonal context. The sample consisted of 155 male and female native English-speaking Canadian undergraduates. Results supported Orvis et al.’s (1975) predictions for the person and stimulus attributions. There was less support for the circumstance attributions, and the least amount of support was found for the person x stimulus attribution. Discussion focused on design and analysis limitations, applications of the current study to other realms in emotion, and the need for a more extensive examination of other emotions.
Research on temporal discounting has primarily used hypothetical monetary rewards. For the colleg... more Research on temporal discounting has primarily used hypothetical monetary rewards. For the college student population, course points serve as a powerful generalized reinforcer, much like money does for the general population. To date, few studies have examined the degree to which students discount points within a course. Silva and Gross (2004, Study 2) found that higher performing students were less likely to discount extra credit points than middle or lower performing students. However, this study focused on the effort to earn the reinforcer, not on the delay of the reinforcer. Williams and colleagues (2007) developed a computerized procedure to examine the rate at which students discounted extra credit points. They found that the majority of students did not discount course points and those who did showed rather shallow discounting curves. They posited that these results may have been due to a magnitude effect (10 course points for a maximum of 55 minutes) or too short of delay periods (ranging from 1 minute to 10 minutes). Further, participants might have been more willing to engage in self-control behaviors on the longer time delays due to prior commitment on the shorter time delays. The current study was designed to address some of the aforementioned limitations in Williams et al. (2007) by lengthening the delays, counterbalancing the presentation order, and looking for evidence of a magnitude effect.
Previous studies have shown that disgust sensitivity (DS), right-wing authoritarianism (RWA), rel... more Previous studies have shown that disgust sensitivity (DS), right-wing authoritarianism (RWA), religious fundamentalism (RF), and belief in a just world (BJW) are associated with social conservatism and religious questioning (QUEST) is associated with social liberalism. The current study was designed determine whether these personality traits would predict who the student planned to vote for in the 2016 U.S. Presidential primary. Consistent with previous research, DS, RWA, RF, and BJW were positively correlated with social conservatism and QUEST was negatively correlated. Students who were highest on RWA, RF, and lowest on QUEST were most likely to vote for one of the Republican candidates. In contrast, students who were at the opposite end of the scales were most likely to vote for Bernie Sanders, with Hillary Clinton supporters in the middle. There were no differences in DS and few differences in BJW with regard to preferred candidate.
Uploads
Papers by Heidi Dempsey