Two hundred and twenty eight Canadian undergraduates read one of two crime vignettes that varied in how justified a murder seemed to be, then made capital (death penalty), capital-related (method of execution), and non capital (prison term, granting of parole, waiting period before parole) sentencing recommendations. Compared to people with a negative attitude toward capital punishment, people with a positive attitude were more severe for both the capital and non capital judgments. They were also harsher for the less than for the more justifiable crime, whereas people with a negative attitude gave similar judgments in each case. Women were generally harsher than men. It is suggested that members of juries chosen to be death-qualified may be biased in any non capital recommendations and that attitude toward capital punishment may reflect other personality traits and values that determine how generally harsh a person's recommended punishment will be.
When a defendant is found guilty, the judgment and subsequent sentence should be based on evidential factors such as crime seriousness and the degree of criminal responsibility (Mazzella & Feingold, 1994). That is, using legally-admissable evidence, punishment should be harsher for crimes that are more serious than those that are less serious, and for people who planned the crime and carried it out intentionally than for those who acted less deliberately. Severity of punishment might also depend on the circumstances in which the crime was committed, particularly if the defendant is judged to have had some justification for committing the crime. Other extralegal factors may also influence offender treatment. Some of these, such as the employment status or sex of the defendant, may be appropriate on utilitarian grounds because they are thought to predict recidivism (Gebotys & Roberts, 1987). For example, an employed person has a lower risk of re-offending than an unemployed person and so may be given a lighter sentence than an unemployed person. However, other factors, such as defendant attractiveness (Mazzella & Feingold, 1994), represent bias and are unacceptable. For example, people who are more attractive have been judged more leniently than people who are less attractive (Mazzella & Feingold, 1994).
These factors refer to the offender, but some refer to the judge and jury, who are often permitted considerable discretion when deciding punishment (Wheeler, Weisburd & Bode, 1982). One of these personal characteristics is attitude toward capital punishment, which may be an acceptable or an unacceptable extralegal factor, depending on the nature of the sentence. For example, as long as a death penalty recommendation follows the law, courts in the United States accept that jurors' attitude towards capital punishment will be related to their verdict, because the death penalty could never be rendered by people who are not "death qualified" (O'Neil, Patry, & Penrod, 2004). However, it would not be appropriate for attitude towards capital punishment to influence non capital sentences.
Furthermore, it would not be appropriate for any sentencing decision to be related to subject factors such as the sex, personality traits or political values of people making judgments. Unfortunately, there is evidence of this kind of bias. For sex, men are usually harsher than women when recommending the death penalty in a specific case (Honeyman & Ogloff, 1996). Although studies show no relationship between sex and length of prison sentence (McKelvie, Mitchell, Arnot, & Sullivan, 1993; Riedel, 1993), some show that men are harsher (McKelvie, 2002) and others show that women are harsher (McNamara, Vattano, & Viney, 1993).
With regard to personality and values, people who are high in authoritarianism have recommended more severe punishment than those who are low (Gerbasi, Zuckerman, & Reis, 1977), particularly if they perceive themselves to have different attitudes than the defendant (Mitchell & Byrne, 1973). Authoritarianism in the person judging and their attitude similarity with the offender have also been related to their attributions of responsibility and leniency of treatment (Feather, 2002). Furthermore, conservatives have been found to be more punitive than liberals and to be more likely to attribute the defendant's actions to internal than to external causes (Davis, Severy, Kraus, & Whitaker, 1993; Vidmar, 1974). In the latter case, intention, which is normally an evidential factor, becomes an extralegal factor, because the attribution is biased by personal characteristics of the judge.
The main purpose of the present study was to investigate the relationship between attitude toward capital punishment and sentencing severity, which has both practical and theoretical implications. Although it would be expected and acceptable that attitude towards capital punishment would be related to a death penalty recommendation, attitude towards capital punishment should not be related to non capital sentencing. As noted above, jurors in the United States are usually screened in capital cases, and this may also occur in other countries (Mauro, 1992). That is, jurors are not permitted to serve unless they are "death-qualified," which essentially means that they must be willing to vote for a death sentence when that is an option (Horowitz & Seguin, 1986; Mauro, 1992). Of course, this does not mean that they will recommend it, only that they would not exclude it from consideration. The practical implication of the present research is that if attitude towards capital punishment is related in general to sentencing severity, it introduces a source of bias towards more severe non capital recommendations when the jury is selected to be death-qualified (Mauro, 1992). Indeed, it has been found that death-qualified jurors are more likely to convict than people who would be excluded from capital cases (Horowitz & Seguin, 1986; Mauro 1992; although for a dissenting view, see Elliot, 1992).
It was noted above that sentencing decisions have been related to personality traits and political values. The present results are of theoretical interest because it has been suggested that belief in the death penalty reflects these traits and values (Stack, 2000). For example, Eysenck (1954) argued that belief in capital punishment was based on the trait of radicalism/conservatism. In addition, compared to people who oppose the death penalty, those who believe in it are more authoritarian (Ray, 1982; Stack, 2000; Vidmar, 1974), less liberal (Stack, 2000), and less tolerant (Valliant & Oliver, 1997). Like conservatives compared to liberals, they are also more likely than opponents of capital punishment to judge that the act was intended (Goodman-Delahunty, Greene, & Hsiao, 1998).
If attitude toward capital punishment reflects such traits and values, it might be expected that people in favour of capital punishment would be generally harsher in their sentencing judgments than those who are against it. That is, because they favour capital punishment, they would be more likely to recommend the death penalty, and possibly a more severe method of execution, but they might also recommend longer prison terms and be more severe in granting parole. In fact, consistent with this argument, and like death-qualified jurors, people who believe in capital punishment are more likely to convict than those who are against it (Allen, Mabry, & McKelton, 1998). In addition, belief in capital punishment might not only be related to perceived intention, but also to other attributions. For example, people who are in favor of capital punishment may judge that the offender was less justified in their action and less driven to it. Such findings have additional theoretical implications, because it has been proposed that attitude towards capital punishment may have an indirect effect on capital sentencing via its relationship with attributions (O'Neil et al., 2004). That is, people with a stronger belief in capital punishment may evaluate the motives of the accused more harshly than people with a weaker belief, leading them to be more likely to recommend the death penalty.
On the other hand, past research has shown that attitudes are a better predictor of behaviour if the two domains share the same level of specificity (Newcomb, Rabow, & Hernandez, 1992). Belief in the death penalty may be a fairly specific attitude, so that those who are pro would be more likely to recommend capital punishment than those are con, but they might not be more severe in other recommendations pertaining to jail time and parole (and perhaps method of execution), and they might not differ in their attributions.
This review shows that people with a more positive attitude toward capital punishment and people who are death qualified are more likely to convict an offender and to perceive that their action was intended than people with a more negative attitude and people who are excludable. The present experiment extends this work by examining the relationship between attitude toward capital punishment and three kinds of sentencing recommendation: capital (death penalty), capital-related (execution method), and non capital (prison term, granting of parole, waiting period before parole). The relationship between attitude toward capital punishment and three kinds of attribution (justification, being driven to act, intent) was also investigated.
Participants began by completing a questionnaire that measured their attitude toward capital punishment. Although there has been some disagreement over whether attitudes should be defined as composed of three components (cognition, affect, behavioral tendency) or only one (affect) (Brigham, 1991, p. 134), the present instrument consisted of statements about belief in capital punishment. As such, it emphasized the cognitive component of the attitude. Participants then read one of two cases in which the accused had already been found guilty of a planned murder. In both conditions, the defendant killed a man whom he felt was responsible for his brother's death, but they differed according to the circumstances under which the act occurred. In one case (killing circumstances), which was expected to be perceived as more justified, the murder victim was a hit man who had himself been accused of killing the brother. In the other (job circumstances), the defendant's brother died suddenly after discovering that his application for a job had been unsuccessful. The murder victim was the person who obtained the position over the defendant's brother. After reading the crime scenarios, participants gave sentencing recommendations then their attributions.
Because the defendant had been found guilty of the murder, it was expected that attitude toward capital punishment would be directly related to the likelihood that the death penalty was recommended. However, the relationship might be stronger in the job circumstances than in the killing circumstances. In the first case, where the offender would be perceived as having very little or no justification for his action, people who had a negative attitude to capital punishment would not recommend it, but people who had a positive attitude would be quite likely to do so. In the second case, where the offender would be perceived as more justified in his action, people with a negative attitude would again not recommend it, but people with a positive might be only slightly more likely to do so.
However, the major question here was whether attitude toward capital punishment would also be related to the capital-related and to the non capital recommendations for punishment and to attributions concerning the criminal act. If people with a positive attitude toward capital punishment are generally harsher than those with a negative attitude, the capital-related and non capital recommendations would follow the same pattern as the capital recommendations. In addition, although the scenarios described both murders as deliberate, which implies that perceived intention would be very high and similar for all participants, people with a positive attitude towards capital punishment might be more likely to perceive the act as intended than people with a negative attitude. Finally, people with a positive attitude might be less likely to perceive the offender as justified or driven to commit the act than those with a negative attitude.
A secondary interest in the present study was sex of participant. In view of the research cited above, it was expected that men would have a more positive attitude toward capital punishment than women and would be more likely to render a death sentence. However, no prediction was made concerning sex differences in the capital-related or non capital judgments, or in the attributions.
Two samples of Canadian undergraduates were employed to develop a severity scale for the different methods of capital punishment. One sample of 64 introductory psychology students (46 women, 18 men) rated and ranked six execution procedures for severity. To check whether these judgments could be replicated by people who had also read one of the two crime scenarios, a second sample of 149 undergraduates (84 women, 65 men) from a variety of disciplines rated the execution methods. They were allocated randomly to the two crime circumstances conditions (killing, job), with proportional matching for sex.
The main part of the experiment was conducted with two additional samples of Canadian university undergraduates (55 women, 52 men; 58 women, 63 men) from a variety of disciplines. Participants were allocated at random to the two crime circumstances conditions (killing, job), with matching for sex. However, within each of these four crime circumstances/sex conditions, participants were classified as having either a positive or a negative attitude toward capital punishment.
Materials and Procedure
Scale for Severity of Execution Methods. The first sample of participants, who rated the execution methods for severity, were told that a judge had imposed the death sentence on a man found guilty of murder. Because different methods could be seen to involve different amounts of punishment, they were asked to rate the severity of six methods (electric chair, firing squad, gas chamber, guillotine, hanging, and lethal injection) on a labeled 5-point scale from 1 = not severe at all to 5 = extremely severe. Methods were presented to each rater in one of four different orders. After completing these ratings, participants then ranked the six methods from 1 = most severe to 6 = least severe. Although they could give the same rating to more than one method, they had to give different ranks to each method.
The six methods were chosen because they are likely to vary in the degree to which people regard them as severe and painful (Hillman, 1993). In addition they are all used in some countries that retain the death penalty (Pratarelli & Bishop, 1998-99). Firing squad is most popular, (96 countries), with hanging not far behind (78 countries). Although the guillotine is not used today, six countries permit beheading, and one country uses electrocution, gassing and lethal injection. Stoning is also on the books in seven countries, but was omitted from the first sample by mistake. However, it was included in the list of methods given to the second sample, who read the crime scenarios and made sentencing recommendations before rating the execution methods. This sample did not rank them. Although these participants also made crime judgments, only their execution method ratings are presented here. These ratings are particularly relevant to the main part of the study because they were obtained by people who also made crime judgments.
Questionnaire on Capital Punishment. The Questionnaire on Capital Punishment (QCP) was slightly different in the two samples of experimental participants. In the first group, it consisted of 18 items that were rated on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree) and was similar to a previous questionnaire (McKelvie, 1983; McKelvie & Daoussis, 1982) that was itself very closely based on a 24-item Attitude to Capital Punishment scale created by Thurstone in the 1930s (Peterson & Thurstone, 1933; Shaw & Wright, 1967, p. 161) using the method of equal-appearing intervals.
Internal consistency reliability estimates for the scale range from .59 to .88 and Thurstone himself reports test-retest reliability to be .71, which Shaw and Wright (1967, p. 161) judge as adequate. With a sample of undergraduates, York (1966) found a correlation of .98 between the scale values obtained by Thurstone and ratings of favourableness of items towards capital punishment, showing that the meaning of the items was stable over time. Moore (1975) found that students who voted in favor of a proposition to bring back the death penalty in California had higher scores on the Thurstone scale than students who voted against, supporting its validity.
In the present QCP, a previous neutral item was replaced by one that was favourable to capital punishment, so that 9 items were favourable and 9 items were unfavourable. In addition, rather than just checking the items with which they agreed, which was Thurstone's procedure, participants responded to each item on a labeled scale from 1 = strongly disagree to 3 = neutral to 5 = strongly agree. Scoring was reversed on unfavourable items so that the total score, which ranged from 18 to 90 (midpoint = 54), measured how favourable the person's attitude was.
In the second group, the QCP was modified again by increasing the number of items to 20 (10 favourable, 10 unfavourable) so that the total score was 20 to 100 (midpoint = 60). The total score for each participant was divided by the number of items in the questionnaire to give a mean rating from 1 to 5. In this way, the questionnaire scores were comparable for the two samples. Examples of items are "Capital punishment has never been effective in preventing crime" and "We must have capital punishment for some crimes."
After completing the QCP, the two groups of experimental participants imagined that they were a member of a jury at a trial in which a man had been found guilty of murder. They were told to read a transcript of the case with a view of subsequently sentencing the guilty party. In approximately 200 words, it described how Charles Weaver, an accountant, was on his way home when he was openly accosted and killed by the defendant, Henry Ellis, who shot Weaver twice in the head. Psychiatrists testified that Ellis was angry when he shot Weaver, but was not suffering from any mental disorder.
In the killing circumstances condition, it was stated that Weaver, the dead man, was really a hit man for the mob, and had been tried for the murder of Henry Ellis' brother John, who was an innocent victim of mistaken identity. However, the case had been dismissed because of insufficient evidence. Ellis admitted that he was avenging his brother's death, which he blamed on Weaver.
In the job circumstances condition, it was stated that Weaver, the dead man, had recently been fairly appointed to an accountancy position. Another candidate was Ellis' brother John. When John Ellis learned that he had been passed over, he became upset and died of a sudden heart attack. As in the killing circumstances condition, Ellis admitted that he was avenging his brother's death, which he blamed on Weaver.
Referring back to the transcript if they wished, participants then answered eight questions in this order: length of prison sentence from 5 to 99 years (prison term), recommendation for parole from 1 (definitely no) to 6 (definitely yes) (parole), number of years to be served before being paroled (parole waiting time), recommendation for the death penalty from 1 (definitely no) to 6 (definitely yes) (death penalty), execution method from seven choices, extent to which Ellis was driven to his action from 1 (not at all) to 7 (very strongly), extent to which Ellis was justified in his action from 1 (not at all) to 7 (very strongly), and extent to which Ellis intended his action from 1 (definitely no) to 6 (definitely yes). Participants in the second sample also stated their reason for choosing the execution method. This question was mistakenly omitted in the first sample.
Participants in each experimental crime circumstances group were classified as having a positive or a negative attitude toward capital punishment on the basis of whether they scored above or below the absolute midpoint of the attitude scale. This corresponded to a mean rating of 3 (= neutral). In the first sample, 64 people (30 women, 34 men) fell into the positive category and 43 people (25 women, 18 men) fell into the negative category. In the second sample, 65 people (23 women, 42 men) fell into the positive category and 56 people (35 women, 21 men) fell into the negative category.
Scoring for most of the dependent variables was simply the numerical response. However, because length of waiting period before parole was made after the recommended sentence was given, a stricter measure of parole severity was calculated: the waiting period divided by the sentence recommendation (proportional parole). This kind of ratio score has also been employed in previous archival (Champion, 1987) and experimental (Bergeron & McKelvie, 2004; McKelvie & Bergeron, 2003) research. Execution method was scored using the severity rating that each participant gave for the chosen method. As described below, the prison term and parole scores were combined to create a joint score for non capital judgments.
A scale was constructed to score the severity of the execution method chosen by the experimental participants. A 2 X 6 (Sex of Rater X Method of Execution) mixed ANOVA with sex of participant as the between-groups factor and method as the within-groups factor was conducted on the severity ratings from the first sample of raters. The only significant effect was method, F(5, 330) = 14.29, p < .001. Post hoc Newman-Keuls tests showed that lethal injection was rated as less severe than the other methods, which did not differ from each other. The mean scores in Table 1 indicate that even the lowest rating of 3.2 for lethal injection was close to scale point 3 that was labeled "severe." The other methods received mean ratings between 3.9 to 4.4, which were close to 4 ("very severe"). A Friedman analysis of variance on the ranks (see Table 1) was also significant (p < .001), with lethal injection as the least severe, gas chamber the next most severe and the other four methods the most severe.
Each participant's rankings were then reversed (1 became 6, 6 became 1 etc.) and summed with their ratings for each method to give a joint rating/ranking scale from 2 (least severe) to 11 (most severe). In a mixed ANOVA of these scores, only method was significant, F(5, 310) = 21.94, p < .001. Post hoc Newman-Keuls tests showed that lethal injection was perceived as less severe than gas chamber, and that gas chamber was perceived as less severe than the other methods, which did not differ (see Table 1).
For the ratings given by the second sample of participants who had read the crime scenarios (Table 1), a 2 X 2 X 7 (Crime Circumstances X Sex of Rater X Method of Execution) ANOVA gave only a significant effect of method of execution, F(6, 708) = 60.37, p < .001. Newman-Keuls tests again indicated that lethal injection was rated as less severe than gas chamber, which in turn was rated as less severe than the other methods. However, stoning, which was not included in the previous choices, was rated as more severe than the electric chair, hanging, guillotine and firing squad, which were not rated differently. Although the mean lethal injection rating of 2.7 was slightly lower than the value of 3.2 from the first group of raters, the means for electric chair, hanging, guillotine and firing squad were again close to 4. This replicates the previous results. The highest mean rating for stoning was 4.6, which is fairly close to 5 ("extremely severe").
Because the ratings of the second sample confirmed those of the first one, they were used to scale the severity of the six methods. The scale value for each method was its mean rating from this sample (last row of Table 1). Because others may use this scale to measure how people vary in their ratings of severity to execution methods in general, alpha was calculated as an estimate of internal consistency and found to be .78.
Because participants in the two crime circumstances conditions were classified as having a positive or a negative attitude towards capital punishment, it was important to check that their scores were matched in the two conditions. It was also important to check that they were matched for the men and women who were positive and negative in each condition. In addition, because the QCP was slightly different in the two samples, sample was included as a variable. In the 2 X 2 X 2 X 2 (Crime Circumstances X Attitude Toward Capital Punishment X Sex of Participant X Sample) factorial ANOVA that was conducted on mean QCP scores, there were two significant effects. People with a positive attitude towards capital punishment had a higher mean score than those who were negative, F(1, 212) = 653.62, p < .001, and scores were higher in the first sample (18 items) than in the second sample (20 items), F(1, 212) = 7.49, p < .01. However, none of the interactions approached significance (ps > .18). These results show that people with a positive attitude can be meaningfully compared to people with a negative attitude for both men and women in the two crime conditions (see Table 2).
Combining data over both experimental conditions, ratings for the extent to which the defendant was driven to commit the offence were significantly correlated with ratings of justification in each sample, r = .411, .308, p < .01, respectively. However, ratings of intention were not significantly related to ratings of being driven (r = .124, 0.020) or to ratings of justification (r = .089, .046).
Because the attributions were partly related to each other, they were examined together with a 2 X 2 X 2 X 2 (Crime Circumstances X Attitude Toward Capital Punishment X Gender X Sample) MANOVA with three dependent variables (justification, being driven, intention). Because there were no main effects or interactions involving gender or sample (ps > .09), another 2 X 2 (Crime Circumstances X Attitude Towards Capital Punishment) MANOVA collapsed over these two variables was conducted. In addition, because the independent variables were expected to be related to more than one dependent variable, Pillai's criterion was used (Tabachnik & Fidell, 1996). Two effects were significant: crime circumstances, F(3, 221) = 43.02, p < .001, and attitude towards capital punishment, F(3, 221) = 3.61, p < .05.
To find out where these effects occurred, the univariate analyses were inspected (see Table 3). For ratings of the extent to which the defendant was driven to commit the crime, the effect of crime circumstances was significant, F(1, 223) = 60.66, p < .001. The defendant was perceived as more driven in the killing than job circumstances. For ratings of justification, the effect of crime circumstances, F(1, 223) = 83.60, p < .001, and of attitude towards capital punishment, F(1, 211) = 6.43, p< .02, were significant. The defendant was perceived as more justified in the killing than job circumstances, and people who had a positive attitude towards capital punishment rated the offender as more justified than people who had a negative attitude. For intention, the main effect of attitude towards capital punishment was significant, F(1, 223) = 3.93, p < .05. Intention was rated higher by people with a positive attitude than by people with a negative attitude.
Because these univariate tests are post hoc tests, it has been suggested that they should be corrected for Type I error using the Bonferroni adjustment. In this case, for three dependent variables, alpha would be set at .017 (Tabachnik & Fidell, 1996). All of the above effects except intention were significant with this guideline.
Although most researchers interpret the results of MANOVA with the univariate tests with the conventional uncorrected level of alpha, these tests assume that the dependent variables are uncorrelated. A stricter approach is the stepdown procedure, in which dependent variables are analyzed in sequence. The theoretically most important one is analyzed first with ANOVA, and the others are analyzed with ANCOVAS in which the previous ones are entered as covariates (Tabachnik & Fidell, 1996). This method pinpoints which dependent variables contribute directly to the multivariate effect.
Because justification was the most important rating, it was examined first. Referring to the analysis above, the effect of crime circumstances (p < .001) and of attitude toward capital punishment (p = .012) were significant at the .017 level.
Using the stepdown procedure, ratings of the extent to which the defendant was driven to commit the act were analyzed with a 2 X 2 (Crime Circumstances X Attitude Toward Capital Punishment) ANCOVA, with justification as the covariate. The effect of crime circumstances was again significant, F(1, 222) = 31.68, p < .001. Intention was not analyzed with the stepdown procedure because it was not correlated with justification or being driven.
For each participant, a summed score was calculated with the four non capital punishment measures (prison term, granting of parole, waiting period before parole, proportional parole). Each raw score was converted to a z-score, which were then averaged. For likelihood of recommending parole, the score was reversed because a higher rating indicated a less severe judgment. For the non capital joint score, internal consistency as represented by alpha was .627, which is acceptable for research purposes given that it was based on only four items.
The death penalty recommendation was retained as a separate variable because it referred directly to capital punishment. Choice of execution method was also kept separate because it reflected harshness of punishment but was also distinct from the other non capital sentencing judgments. However, to keep these two scores comparable with the non capital joint score, they were also converted to z-scores.
In each experimental sample, correlations showed that participants who were more likely to recommend the death penalty chose a more severe execution method (.219, p < .05, .265, p < .01; .247, p < .001 for both samples together) and made harsher non capital judgments (.474, p < .001, .530, p < .001; .497, p < .001 for both samples together). In addition, people who chose a more severe method also made harsher non capital judgments in the second sample (.308, p < .005), but not in the first sample (-.033, p >.700), leading to a significant correlation of .145, p < .05 for both samples together.
A 2 X 2 X 2 X 2 (Crime Circumstances X Attitude Towards Capital Punishment X Sex X Sample) MANOVA was conducted on the capital, capital-related, and non capital scores as the dependent variables. Sample was kept as an independent variable because the QCP was slightly different (18 items, 20 items) in the two cases. A number of factors were significant, but one involved sample: the four-way interaction among all variables, F(3, 200) = 3.19, p < .03. The univariate tests showed that this effect was significant only on the death penalty variable, F(1, 217) = 9.40, p < .003. Inspection of the mean scores showed that the death penalty was rated as more likely by people with a positive attitude than those with a negative attitude in all conditions except one, where the means were the same: men in the first sample in the killing condition. Because there were no consistent differences in the pattern of mean scores for the QCP for sample, it was dropped from the analysis, which was re-run with a 2 X 2 X 2 (Crime Circumstances X Attitude Toward Capital Punishment X Sex) MANOVA. The mean z-scores are shown in Table 4 and the mean raw scores on all judgments are shown in Table 5. Although the latter were not analyzed separately, they show the numbers on which the z-scores were based.
In this analysis, four effects were significant: crime circumstances, F(3, 208) = 8.87, p < .001, attitude toward capital punishment, F(3, 208) = 32.12, p < .001, the interaction between crime circumstances and attitude toward capital punishment, F(3. 208) = 3.44, p < .02, and sex, F(3, 208) = 2.87, p < .04.
As with attributions, the univariate analyses were interpreted with alpha set at .05, the usual practice, but also with the Bonferroni correction. Because there were three dependent variables, alpha was .017.
For the death penalty recommendation, the univariate analysis showed that there were significant effects of crime circumstances, F(1, 210) = 15.32, p < 001, attitude toward capital punishment, F(1, 210) = 94.56, p < .001, and the interaction between them, F(1, 210) = 8.70, p < 003. The death penalty was more likely to be recommended in the job condition than in the killing condition (standardized effect size d = 0.55), and was more likely to be recommended by people with a positive attitude towards capital punishment than those with a negative attitude (d = 1.37). Post hoc t-tests showed that people with a positive attitude toward capital punishment were more likely to recommend the death penalty than people with a negative attitude in both the killing circumstances, t(210) = 4.99, p < .001, and the job circumstances, t(210) = 9.11, p < .001. However, the standardized effect size was greater in the job circumstances than in the killing circumstances (d = 1.81, 0.95 respectively). In addition, although people with a positive attitude were more likely to recommend the death penalty in the job circumstances than in the killing circumstances, t(210) = 5.43, p < .001, d = 0.97, this comparison was not significant for people with a negative attitude, t(210) = 0.65, p > .10, d = 0.14. All these significant effects held with alpha at .05 and at .017.
For the method of execution, there was a significant effect of crime circumstances, F(1, 210) = 5.24, p < .03, d = 0.33, and the effect of attitude towards capital punishment was almost significant, F(1, 210) = 3.56, p < .07, d = 0.27. A more severe method was recommended in the job circumstances than in the killing circumstances and by people with a positive attitude towards capital punishment than those with a negative attitude. However, neither effect was significant with the Bonferroni correction at .017.
For the combined score for non capital judgments, there was a significant effect of crime circumstances, F(1, 210) = 19.15, p < .001, d = 0.62, attitude towards capital punishment, F(1, 210) = 7.69, p < .01, d = 0.39, and of the interaction between crime circumstances and attitude toward capital punishment, F(1, 210) = 5.88, p = .016. These effects were significant with the conventional level of alpha at .05 and with the Bonferroni correction. Punishment was more severe in the job condition than in the killing condition and was more severe for people with a positive attitude towards capital punishment than for those with a negative attitude. Post hoc t-tests showed that people with a positive attitude toward capital punishment gave more severe recommendations than people with a negative attitude in the job circumstances, t(210) = 3.73, p < .001, d = 0.74, but not in the killing circumstances, t(210) = 0.26, p > .10, d = 0.05. In addition, people with a positive attitude t(210) = 4.36, p < .001, and people with a negative attitude t(210) = 2.21, p < .05, made more severe judgments in the job circumstances than in the killing circumstances, but the effect size was greater in the first case (d = 0.96, 0.28 respectively). With exception of the last comparison, all the significant effects held with the Bonferroni alpha at .017. Finally, the effect of sex approached significance at the conventional level, F(1, 210) = 3.86, p < .06, d = 0.46, but not with the Bonferroni correction. Women rendered harsher judgments than men.
ANCOVA was conducted on the method of execution, with the death penalty score as a covariate. None of the effects were significant. For non capital judgments, with both death penalty and execution method as covariates, there were significant effects of crime circumstances, F(1, 208) = 7.94, p < .006, sex, F(1, 208) = 5.90, p = .016, and the interaction between crime circumstances and sex, F(1, 208) = 5.12, p < .03. However, only main effects of crime circumstances and sex were significant with the Bonferroni correction. Judgments were harsher in the job condition than in the killing condition and women rendered harsher judgments than men. The same results were obtained when only death penalty was entered as a covariate. Note that these covariance analyses are conservative because although the death penalty had priority in terms of theoretical importance, which implies that it should be entered, participants made non capital judgments before their capital judgment. This means that their non capital judgments were made without reference to the capital judgment and would not have been influenced by it.
A 2 X 2 X 2 (Crime Circumstances X Attitude Toward Capital Punishment X Sex) MANCOVA was conducted on the capital, capital-related, and non capital judgments with scores on the three attributions as covariates. There were significant effects of crime circumstances, F(3, 204) = 2.85, p < .04, attitude toward capital punishment, F(3, 204) = 30.84, p < .001, and the interaction between crime circumstances and attitude toward capital punishment, F(3, 204) = 2.72, p < .05.
The frequency with which each execution method was chosen did not vary among the conditions, and ranked as follows: lethal injection (45.6%), electric chair (11.9%), firing squad and gas chamber (10%), guillotine (2.5%), hanging (1.3%), and stoning (0.6%). Reasons for choosing each method were classified as humanitarian (minimizing suffering for the criminal), punishment (making the criminal suffer) or utilitarian (practical), and ranked as follows. For lethal injection: humanitarian (100%); for all other methods: humanitarian (50%), punishment (45.8%), and utilitarian (4.2%).
As predicted, people with a positive attitude towards capital punishment were more likely to recommend the death penalty than those with a negative attitude, and this effect was stronger in the job circumstances where the victim was fairly appointed to a professional position than in the killing circumstances where the victim was a former hit man who himself had been tried for murder. The standardized effect sizes were 1.81 and 0.95 respectively, which both exceed Cohen's (1977) standard of 0.80 for large. The reason for expecting that the effect would be stronger in the first case was that the murder of a man who was appointed fairly to a job would be seen as less justified and therefore more punishable than the murder of a man who had himself been accused of a killing. Indeed, participants rated the murder in the job circumstances as less justified than in the killing circumstances, and they also rated that offender as less driven to his action in the job circumstances. Furthermore, participants' capital and non capital recommendations were both harsher in the job circumstances than in the killing circumstances. These results demonstrate that people with a positive attitude toward capital punishment were more likely to recommend the death penalty for a specific crime than people with a negative attitude, and that they took the circumstances into account. In fact, as predicted, people with a negative attitude were unlikely to recommend the death penalty in either scenario, whereas those with a positive attitude were more likely to recommend it than those with a negative attitude in the killing circumstances, and were also more likely to recommend it in the job circumstances than in the killing circumstances. Although attitude toward capital punishment is an extralegal factor, its relationship to the death penalty recommendation is acceptable, because in countries which retain capital punishment, juries are usually death-qualified. (O'Neil et al., 2004). However, any relationship to other kinds of recommendations are not acceptable, and represent a bias in judgment.
The main purpose of the present study was to examine this question by establishing whether attitude toward capital punishment was associated with harshness of punishment in general. For the method of execution, which was not itself a capital judgment but was related to it, people with a positive attitude toward capital punishment chose a slightly more severe method than those with a negative attitude. The effect size was 0.27, which is close to Cohen's (1977) standard for small. However, this difference was only marginally significant by the conventional standard and not significant with the Bonferroni correction or when analyzed with death penalty entered as a covariate. Overall, the strong relationship between attitude toward capital punishment and death penalty recommendation did not generalize to the method of execution. One reason for this may be that most people chose lethal injection for humanitarian reasons, which is consistent with the low mean score of 3 (minimum 2.7) for severity of method chosen (see Table 5). Notably, for participants who chose another method, both humanitarian and punishment reasons were cited, but the number of participants may have been too small to show any effect of attitude towards capital punishment. At the same time, the present finding is also consistent with a report that belief in capital punishment was not related to perception of pain from different methods of execution (Pratarelli & Bishop, 1998-99).
However, for the non capital recommendations, which included prison time, granting of parole, and waiting time before parole, the results were similar to those for the capital recommendation. Although they were not significant with both death penalty and execution methods as covariates, showing that the non capital recommendations did not contribute independently to the overall multivariate effect, the non capital judgments were made prior to and independently of the capital judgment. Most important, it was found that the effects of attitude towards capital punishment and the interaction between crime circumstances and attitude toward capital punishment were significant on the univariate tests with the strict Bonferroni correction. People with a positive attitude toward capital punishment were harsher than people with a negative attitude, and this effect was stronger in the job circumstances than in the killing circumstances, where the effect sizes were 0.74 (large) and 0.05 respectively. These results show that the relationship between attitude toward capital punishment and capital recommendations generalized to non capital recommendations. They suggest that people with a positive attitude toward capital punishment would be likely to render harsher prison terms and parole recommendations than those with a negative attitude. Because this would not be acceptable under the law, it demonstrates an extralegal factor that is a source of bias in judgment (Mauro, 1992). Practically, the results imply that if a death-qualified jury did not make a capital recommendation, it would be likely to recommend harsher non capital punishment than a jury that was not death-qualified.
These findings are also consistent with previous research demonstrating that people who believe in capital punishment and jurors who are death-qualified are more likely to convict an accused person than people who are opposed and not death qualified (Allen, et al., 1998; Horowitz & Seguin, 1986; Mauro 1992). However, they extend past research to the second stage of jury deliberations where sentencing occurs. In most capital court cases, the same jurors reach a verdict then, at a later time, make a sentencing recommendation (a bifurcated system, Horowitz & Seguin, 1986). The results are relevant to the suggestion that different juries be employed in the two phases (Horowitz & Seguin, 1986). In addition, they imply that attitude toward capital punishment may also predict sentences in non capital and perhaps even non criminal cases. People with a positive attitude toward capital punishment might be harsher than those with a negative attitude in civil lawsuits and in other situations involving punishment (e.g., disciplining children). (1)
The relationship between attitude toward capital punishment and attributions was also investigated because it was speculated that people with a positive attitude might judge the offender's motives more harshly than people with a negative attitude. That is, people with a positive attitude might see the offence as less justified and the offender more driven to his action than people with a negative attitude. As in previous research, attitude toward capital punishment might also be related to perceived intention (Goodman-Delahunty, et al., 1998). It was found that people with a positive attitude toward capital punishment did perceive the murder as more intended than people with a negative attitude, although this effect was not significant with the Bonferroni correction. However, there was no relationship between attitude toward capital punishment and the offender being driven. Furthermore, people with a positive attitude perceived the offender as more justified than people with a negative attitude. This was a surprise, because it had been expected that people with a positive attitude would perceive the offender as less justified. Despite the present perception, people with a positive attitude were more punitive than those with a negative attitude. This pattern of results contrasts with the effect of crime on perceived justification and on punishment. The offender was perceived as more justified in the killing circumstances than in the job circumstances and the capital and non capital recommendations were less harsh in the killing circumstances than in the job circumstances.
Although it is not clear why perceived justification was associated differently with punishment for the crime and for attitude toward capital punishment, these results, together with the weak or zero-order relationship between attitude toward capital punishment and the other attributions, do not support O'Neil et al.'s (2004) theoretical proposal that the relationship between attitude toward capital punishment and capital and non capital sentencing recommendation is mediated by perceived motives. This conclusion is supported by the analysis of sentencing recommendations in which attributions were included as covariates. This did not alter the significant effects of attitude toward capital punishment or of the interaction between crime circumstances and attitude toward capital punishment.
What, then, is the basis of the present relationship between attitude toward capital punishment and sentencing? The finding that people with a positive attitude to capital punishment were harsher in their capital recommendations than people with a negative attitude is consistent with the theoretical suggestion that attitudes are more likely to predict behavior if they share the same level of specificity (Newcomb, et al., 1992). However, this does not account for the relationship between attitude toward capital punishment and non capital recommendations, which is more consistent with the proposal that belief in the death penalty reflects other personality traits or values (Eysenck, 1954; Stack, 2000). Although these factors were not measured here, it has been found that attitude toward capital punishment is associated with authoritarianism and conservatism (Ray, 1982; Stack, 2000; Vidmar, 1974), and that both of them are associated with harsher punishment (Davis, et al., 1993; Vidmar, 1974). It would be interesting to investigate these variables simultaneously in future research.
Although of secondary interest, sex of the person judging was also investigated as a possible extralegal factor in sentencing recommendations. Sex of participant was significant in the multivariate analysis, suggesting that women were generally harsher than men. In the univariate analysis, it was not significant for the capital recommendation and was only marginally significant for the non capital recommendations. However, with the capital recommendation as a covariate, women were found to deliver harsher non capital recommendations than men. Together with the fact that men and women did not score significantly differently on the Questionnaire on Capital Punishment (QCP), these results contradict previous research showing that men are more likely to recommend the death penalty than women (Honeyman & Ogloff, 1996). The present finding that women were harsher on non capital recommendations has been reported previously (McNamara, et al., 1993), but others have found than men were harsher (McKelvie, 2002) or that there was no sex difference (McKelvie, et al., 1993; Riedel, 1993). Clearly, more work needs to be done to determine the conditions under which sex differences do and do not occur.
An important component of this study was the measurement of attitude toward capital punishment. Because this attitude has often been measured with fewer than five items and sometimes only one (O'Neil et al., 2004), the 18 or 20 items used here in the QCP are a considerable improvement over past practice. Although these items focus on the cognitive aspect of the attitude to the exclusion of the affective aspect, they covered the major reasons why people believed in the death penalty: deterrence and punishment (retribution) (Honeyman & Ogloff, 1996; Vidmar, 1974). In addition, the psychometric properties of the scale are acceptable for research purposes. Although the number of items was slightly different in the two samples of participants, the mean QCP scores were matched in the different conditions for people classified as having a positive or a negative attitude, and the relationship between attitude toward capital punishment and sentencing recommendations was not affected by this discrepancy.
Although the QCP seems acceptable as a psychometric instrument, attitude toward capital punishment was investigated by splitting scores at the midpoint of the scale and classifying people above the midpoint as having a positive attitude and people below the midpoint as having a negative attitude. Although this strategy loses information, it is consistent with the practice of splitting people at the median of a scale when a subject variable is being studied. However, the median was not chosen here because the midpoint was a more logical cutting point for dividing people into those for and against the death penalty.
It might be objected that most of the methods of execution that were offered as choices are not used in the U.S. However, some are used in other countries, and they were included to create the variation in perceived severity to investigate just how severe people might be. Indeed, other studies in which methods of execution have been investigated include most of the choices offered here (Hillman, 1993; Pratarelli & Bishop, 1998-99). Finally, the scale of severity that was developed was strengthened by the fact that raters were presented with the methods of execution in four different orders.
Although it was hypothesized that perceived intention might vary as a function of attitude toward capital punishment, it was clear in both crime scenarios that the offender planned the murder and carried it out deliberately. The fact that the mean rating of intent was very high (5.5 out of 6) and did not vary according to crime circumstances demonstrates that participants read the scenarios carefully. It was also found that perceived justification was higher in the killing circumstances than in the job circumstances, which validates the manipulation.
In the present study, Canadian undergraduates read an artificial scenario and made individual judgments about how the guilty party should be punished. In real life, jurors from the general population are presented with more detailed evidence, including arguments for the prosecution and for the defense, and they arrive at joint decisions after discussion and consultation with each other. Moreover, the death penalty has been abolished in Canada, so this is not a judgment that the participants would ever be called upon to make in court.
On the other hand, the results show that previous U.S. demonstrations of a relationship between attitude towards capital punishment and conviction rate can be generalized to sentencing decisions in another population. In addition, the artificial laboratory situation allows variables to be manipulated and controlled so that theories can be tested, and it shows what can happen when certain conditions are met (Mook, 1983). In the present experiment, crime circumstances were experimentally manipulated and extraneous variables were controlled via random assignment. Attitude toward capital punishment was only manipulated as a subject variable, but care was taken to ensure that participants in each condition were matched on their QCP scores. These controls ensure that the interactive effect between attitude towards capital punishment and crime circumstances is real: the smaller effect of attitude on capital and non capital sentencing recommendations in the killing than in the job circumstances did not occur because attitude differences were smaller in the first than in the second scenario.
As expected, attitude toward capital punishment was related to the capital recommendation, but it was also related to the non capital recommendations (prison, parole). There was also evidence that women were harsher than men on their non capital recommendations. Practically, this implies that juries consisting of people in favour of capital punishment (e.g. death-qualified juries) may be biased towards severity if called upon to deliver a non capital sentence. The sex difference obtained here also implies that it would be advisable to have an equal number of women and men. Theoretically, the results are consistent with the view that attitude toward capital punishment reflects other traits or values. In future research on attitude toward capital punishment, measures of these traits and values should also be administered and sex differences should be examined to find out when they do and do not occur. In addition, the relationship between attitude toward capital punishment and non capital crimes should be examined to find out the extent to which believers in the death penalty are generally punitive.
Author Note: This work was supported in part by a grant from the Bishop's University Senate Research Committee.
Allen, M., Mabry, E., & McKelton, D. M. (1998). Impact of juror attitudes about the death penalty on juror evaluations of guilt and punishment: A meta-analysis. Law and Human Behavior, 22, 715-731.
Bergeron, C. E., & McKelvie, S. J. (2004). Effects of defendant age on severity of punishment for different crimes. Journal of Social Psychology, 144, 75-90.
Brigham, J. C. (1991). Social psychology, (2nd ed). New York: HarperCollins Publishers Inc.
Champion, D. J. (1987). Elderly felons and sentencing severity: Intergenerational variations in leniency and sentencing trends. Criminal Justice Review, 12, 7-14.
Cohen, J. (1977). Statistical power analysis for the behavioral sciences (Rev. Ed.). New York: Academic Press.
Davis, T. L., Severy, L. J., Kraus, S. J., & Whitaker, J. M. (1993). Predictors of sentencing decisions: The beliefs, personality variables, and demographic factors of juvenile justice personnel. Journal of Applied Social Psychology, 23, 451-477.
Elliot, R. (1992). On the alleged prosecution-proneness of death qualified jurors and juries. In P. Suedfeld and P. E. Tetlock, Psychology and social policy (pp. 255-265). New York: Hemisphere Publishing Corporation.
Eysenck, H. J. (1954). The psychology of politics. London: Methuen & Co.
Feather, N. T. (2002). Reactions to supporters and opponents of uranium mining in relation to status, attitude similarity, and right-wing authoritarianism. Journal of Applied Social Psychology, 32, 1464-1487.
Gebotys, R. J., & Roberts, J. V. (1987). Public views of sentencing: The role of the offender. Canadian Journal of Behavioural Science, 19, 479-488.
Goodman-Delahunty, J., Greene, E., & Hsiao, W. (1998). Construing motive in videotaped killings: The role of jurors' attitudes towards the death penalty. Law and Human Behavior, 22, 257-271.
Hillman, H. (199). The possible pain experienced during executions by different methods. Perception, 22, 745-753.
Honeyman, J. C., & Ogloff, J. R. P. (1996). Capital punishment: Arguments for life and death. Canadian Journal of Behavioral Science, 28, 27-35.
Horowitz, I. A., & Seguin, D. G. (1986). The effects of bifurcation and death qualification on assignment of penalties in capital crimes. Journal of Applied Social Psychology, 16, 165-185.
Mazzella, R., & Feingold, A. (1994). The effects of physical attractiveness, race, socioeconomic status, and gender of defendants and victims on judgments of mock jurors: A meta-analysis. Journal of Applied Social Psychology, 25, 1315-1344.
Mauro, R. (1992). Tipping the scales towards death: The biasing effects of death qualification. In P. Suedfeld and P. E. Tetlock, Psychology and social policy (pp. 243-254). New York: Hemisphere Publishing Corporation.
McKelvie, S. J. (1983). Personality and belief in Capital Punishment: A replication and extension. Personality and Individual Differences, 4, 217-218.
McKelvie, S. J. (2002). Effects of sex of judge and sex of victim on recommended punishment of a male murderer in a mock scenario. Psychological Reports, 91, 533-536.
McKelvie, S. J., & Bergeron, C. E. (2003). Defendants' age and severity of punishment for two crimes in a mock scenario: Further data. Perceptual and Motor Skills, 97, 621-624.
McKelvie, S. J., & Daoussis, L. (1982). Extraversion and attitude toward Capital Punishment. Personality and Individual Differences, 3, 341-342.
Mitchell, H. E., & Byrne, D. (1973). The defendant's dilemma: Effects of jurors' attitudes and authoritarianism on judicial decisions. Journal of Personality and Social Psychology, 25, 123-129.
Mook, D. G. (1983). In defense of external invalidity. American Psychologist, 38, 379-387.
Moore, M, (1975). Attitude toward capital punishment: Scale validation. Psychological Reports, 37, 21-22.
Newcomb, M. D., Rabow, J., & Hernandez, A. C. R. (1992). A cross-national study of nuclear attitudes, normative support, and activist behavior: Additive and interactive effects. Journal of Applied Social Psychology, 22, 780-200.
O'Neil, K. M., Patry, M. W., & Penrod, S. D. (2004). Exploring the effects of attitudes toward the death penalty on capital sentencing verdicts. Psychology, Public Policy, and Law, 10, 443-470.
Peterson, R. C., & Thurstone, L. L. (1933). Motion pictures and the social attitudes of children. New York: The Macmillan Company.
Pratarelli, M. E., & Bishop, J. L. (1998-99). Perceptions of estimated pain experience during execution: Effects of gender and belief in capital punishment. Omega: Journal of Death and Dying, 38, 103-111.
Ray, J. J. (1982). Attitude to the death penalty in South Africa--with some international comparisons. The Journal of Social Psychology, 116, 287-288.
Shaw, M. E., & Wright, J. M. (1967). Scales for the measurement of attitudes. New York: McGraw-Hill Book Company.
Stack, S. (2000). Support for the death penalty: A gender-specific model. Sex Roles, 43, 163-179.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics, 3rd. ed. New York: HarperCollins.
Valliant, P. M., & Oliver, C. L. (1997). Attitude toward capital punishment: A function of leadership style, gender and personality. Social Behavior and Personality, 25, 161-168.
Vidmar, N. (1974). Retributive and utilitarian motives and other correlates of Canadian attitudes toward the death penalty. The Canadian Psychologist, 15, 337-356.
Wheeler, S., Weisburd, D., & Bode, N. (1982). Sentencing the white-collar offender: Rhetoric and reality. American Sociological Review, 47, 641-659.
York, CV. M. (1966). Stability of Thurstone scale values after 35 years. Perceptual and Motor Skills, 23, 628.
(1) I thank an anonymous reviewer for this suggestion.
Stuart J. McKelvie
Author info: Correspondence should be sent to: Dr. Stuart McKelvie, Dept. of Psychology, Bishop's University, 2600 College St., Sherbrooke, J1M0C8, Quebec, Canada.
North American Journal of Psychology, 2006, Vol. 3, No. 8, 567-590.
TABLE 1 Severity Ratings and Ranking for Methods of Execution Group Lethal Gas Injection Chamber Guillotine Hanging Rater Group Ratings 3.2 3.9 4.2 4.2 Ranks 5.4 3.5 3.0 2.9 Joint Score 4.8 6.9 8.1 8.5 Experiment Group Ratings 2.7 3.5 4.0 4.0 Electric Firing Chair Squad Stoning Rater Group Ratings 4.3 4.4 -- Ranks 3.0 2.7 -- Joint Score 8.0 8.6 -- Experiment Group Ratings 3.9 4.2 4.6 Notes. Ratings were made on a scale from 1 = not severe at all to 5 = extremely severe. Ranks ranged from 1 = most severe to 6 = least severe. The joint score ranged from 2 = least severe to 11 = most severe. TABLE 2 Mean Scores on the Questionnaire on Capital Punishment in Each Condition Women Crime Circumstances n M SD Killing Positive Attitude 26 3.74 0.38 Negative Attitude 30 2.18 0.48 Job Positive Attitude 27 3.62 0.27 Negative Attitude 30 2.08 0.49 Men Crime Circumstances n M SD Killing Positive Attitude 36 3.70 0.38 Negative Attitude 24 2.27 0.51 Job Positive Attitude 40 3.62 0.32 Negative Attitude 15 2.22 0.56 Note. Ratings were from 1 = strongly disagree to 5 = strongly agree. TABLE 3 Mean Ratings for Attributions in Each Condition Driven Crime Circumstances n M SD Killing Positive 62 6.02 1.00 Negative 54 5.39 1.45 Job Positive 66 3.89 2.10 Negative 45 4.04 1.89 Justification Intent Crime Circumstances M SD M SD Killing Positive 3.45 1.64 5.66 0.85 Negative 2.76 1.53 5.52 0.69 Job Positive 1.56 1.20 5.56 0.70 Negative 1.33 0.77 5.27 1.05 Note. Ratings were from 1 = not at all to 7 = very strongly for driven and justification, from 1 = definitely no to 6 = definitely yes for intention. TABLE 4 Mean z-scores for Capital (Death Penalty), Capital-Related (Execution Method) and Non Capital Sentencing Judgments in Each Condition Capital Attitude and Crime Circumstances n M SD Killing Positive Attitude 59 0.10 0.97 Negative Attitude 50 -0.67 0.51 Job Positive Attitude 67 0.86 0.83 Negative Attitude 45 -0.57 0.69 Capital-Related Attitude and Crime Circumstances n M SD Killing Positive Attitude 59 -0.06 1.00 Negative Attitude 50 -0.31 0.67 Job Positive Attitude 67 0.30 1.16 Negative Attitude 42 -0.04 0.93 Non Capital Attitude and Crime Circumstances n M SD Killing Positive Attitude 59 -0.22 0.68 Negative Attitude 50 -0.21 0.61 Job Positive Attitude 67 0.41 0.71 Negative Attitude 42 -0.05 0.43 TABLE 5 Mean Scores for All Sentencing Recommendations in Each Condition Crime Waiting Circumstances Sentence Parole Period Killing Positive Attitude n 62 62 60 M 35.88 3.74 17.29 SD 32.60 1.63 14.19 Negative Attitude n 54 54 54 M 36.82 3.56 17.76 SD 32.10 1.46 16.48 Positive Attitude n 67 67 67 M 62.69 2.58 37.49 SD 31.73 1.34 25.30 Negative Attitude n 45 45 45 M 48.37 3.18 22.04 SD 30.79 1.23 12.71 Crime Prop. Death Execution Circumstances Parole Penalty Method Killing Positive Attitude 60 62 61 .62 2.83 3.00 .50 1.58 0.55 Negative Attitude .54 54 50 .57 1.57 2.84 .28 0.82 0.37 Positive Attitude 67 67 67 .61 4.09 3.18 .21 1.36 0.63 Negative Attitude 45 45 42 .55 1.69 2.99 .27 1.10 0.51 Note. Score ranges were 5 to 99 years (sentence), 1 = definitely no, to 6 = definitely yes (parole, death penalty), 2.7 to 4.6 (method). Scores are listed in the order in which they were made.…