of 1
Department Psychology Atkinson Faculty of Liberal and Professional Studies AK/AS/SC/PSYC2022 3.0M Statistical Methods II
Assignment 1 Revised Feb. 4, 2004
Name: Erin C. Ross Student Number: 999999999
2 Gender and Skill Level Influences on Starting Salary Decisions Method and Design This study explores the effect of the skill level and the gender of a hypothetical job applicant on recommendations for entry pay levels. Participants Forty-eight research participants were recruited for this study from an Honours Business Administration programme at a university in a large metropolitan area. All participants had completed at least 60 credits in their programme, including a course in Human Resources. There were 24 male and 24 female participants, who were an average of 21.6 years of age. Method Participants were informed that this was a study exploring hiring decisions, and were told that they would read about a job applicant, rate the candidate on a number of scales and then make recommendations regarding starting salary. They all signed an informed consent document, which informed them that their responses were anonymous and confidential, they could terminate their participation at any time or omit responses to any question, and they would be fully debriefed about the purpose of the study when they were finished participating. Upon arrival at the testing room, research participants were randomly assigned to receive one of six possible hypothetical job applicant descriptions. Once they had finished reading the description, they rated the person on a series of scales, and were asked to make a recommendation for the applicants starting salary. This was presented to participants in a series of 12 ranges, with 1 representing $20,000 - $21,999 and 12
3 representing $42,000-$43,999. Each scale point was a $2000 increase in salary. Following this, participants were fully debriefed and thanked for their participation. The two variables manipulated in this study were gender and skill level of the hypothetical job applicant for a position as a junior accounting clerk. The applicant was described as either John Smythe or Jane Smythe, and was described on the application form as having only a high school accounting course and no experience (low skill level), as having one year of a community college accounting programme and six months experience (moderate skill level) or as having a community college accounting diploma and one year of experience (high skill level). It was expected that the starting salary recommendation would increase as the level of skill increased, and that the male applicant would receive a higher starting salary recommendation than the female. Further, it was expected that the effect of skill level on pay would differ as a function of the gender of the applicant. Statistical hypotheses are as follows: H o : 1 = 2 = H o : 1 = 2 = 3 = H1 : 1 2 H1 : 1 2 3
H o : 11 = 12 = 13 = 21 = 22 = 23 = H1 : 11 12 13 21 22 23
Results For this study, only the salary recommendations were analyzed. Prior to conducting analysis of variance (ANOVA), the assumptions of normality and homogeneity of variance were tested. There is no evidence that these data violate either assumption. The salary ratings were then analysed using a 2 (gender) X 3 (skill level) between-subjects ANOVA. This analysis found significant main effects for gender,
4 F(1,42) = 65.33, p < .05, and skill level, F(2,42) = 130.75, p < .05, as well as a significant interaction of gender and skill level, F(2,42) = 33.08, p < .05. The effects accounted for 30.2%, 63.4% and 30.5% of error, respectively. Males (M = 8.67) received a significantly higher salary recommendation than did females (M = 6.33). This corresponds to roughly $35,000 versus $31,000. Post-hoc tests of means for the main effect of skill level, using Tukeys test, indicate that applicants described as high (M = 9.63) and moderate (M = 8.63) in skill level did not differ significantly salary in recommendations, but both differed significantly from applicants described as low skill level (M = 4.25). Finally, the interaction means were tested using Tukeys test. These means are presented in Table 1. Females and males of low skill level were given salary recommendations that do not differ significantly. For moderate skill levels, males were given significantly higher ratings than were females and the same pattern emerges for high skill level applicants. Within females, those of high skill level were given significantly higher mean salary recommendations than those of low skill level. Within males, those with low skill level were given significantly lower salary recommendations that those of moderate or high skill level.
Discussion
This study explored the effect of skill level and gender of a hypothetical job applicant on recommendations for starting salary. We found that males were given significantly higher starting salary recommendations than females, and that those with a low skill level were given significantly lower salary ratings than those with moderate or
5 high skill. Finally, gender and skill level interacted such that males with high or moderate skill were given significantly higher salary recommendations that were males of low skill or females of moderate or low skill. The magnitude of difference between these salary recommendations was in the area of $5000 per annum, representing a 17% difference in pay. Previous research has indicated that gender plays a role in the evaluation of job applicants as well as in the remuneration of existing employees. Despite the implementation of pay equity programmes in Canada, women continue to get paid less for work of equal value, and the results of this study suggest that this inequity can exist from the point of hiring. Recent research on unintended discrimination suggests that the manner in which this occurs is not through differential evaluation of skills or abilities. Rather, this discrimination operates directly on variables which relate to valuation of the individual. These findings are consistent with this perspective, as they suggest that, at least at moderate skill levels, males receive higher salary recommendations than females. The data are not, however, entirely consistent with this perspective in that males and females of low or high ability did not receive significantly different sala...