Student Life Survey

I recently conducted a survey of DePaul students on behalf of the Department of Student Life to explore their interests in personal leadership development. The survey consisted of five open-ended questions and one set of questions asked through a Likert scale framework. Students were also asked several demographic questions such as their gender. The survey was sent to the entire residence hall student population (approximately 1600); 134 completed surveys were returned. Because of the relatively low rate of return, for the purposes of this content analysis I consider the 134 students to be the whole population, not a sample eligible to be generalized to the whole residence hall population. I chose to focus my analysis on this survey because it consolidates my workload, thus enhancing my in-class and extracurricular performance.

I would like to conduct a two factor content analysis on a portion of the survey results. The first factor to be examined is respondent gender. This variable has two obvious levels, male and female. The second factor stems from the students' answer to the first question on the survey, "Identify one leader that you most admire. What are the primary characteristics of this person that you find most attractive?" The second variable is the gender of the respondent's most admired leader; again, male and female are the two levels of this factor. By examining these variables, one is able to draw correlations between the respondent's gender and that of the person they most admire. Such information may predict the unconscious importance of gender in leadership programming.

The coding scheme of my analysis is rather simple. All respondents who circled "male" as their gender will be counted as "male respondents." Likewise, respondents who circled "female" as their gender will be counted as "female respondents." The gender of the leader a respondent most admires shall be determined by the person's name and pronouns used in reference to the person. The status of "male', and "female" for both factors are determined by the respondent's opinion, not necessarily proven anatomical fact. This coding scheme is exclusive because, in general, a person cannot be both a male and female. It is exhaustive because every person is either male or female. Although I recognize the existence of hermaphrodites or self-proclaimed "asexual" people, factoring in their very small population as an option in this study would skew the results of my analysis.

For the most part, all coding was manifest. The respondent's gender was assessed using the objective rule of placing respondents in one of two categories based on which gender was circled. The second variable was based on the admired leader's name and the pronouns used in reference to the person. Again, names that are known to be exclusively male according to common knowledge were counted as males; the same for women. If no name was given, or the sex of the name was ambiguous, pronouns were examined. Although these two rules proved adequate for my analysis, if a name remained ambiguous, a latent coding scheme would be employed to "go with the researcher's gut feeling." Because I used a manifest coding scheme, my results are very reliable - everyone using his method will get the same answers. Secondly, because I am dealing with such a relatively straight forward factor such as gender, my validity is also good; especially for a manifest coding scheme.

The following observations were made by simply reviewing each of the 134 surveys returned. I kept a tally of male respondents who admire a male leader, female respondents who admire a male leader, male respondents who admire a female leader and female respondents who admire a female leader. I found the following results:

I did not encounter any problems while coding the survey results.

A multiple sample chi-square analysis was performed on the above data to determine its significance. I developed the theoretical values (T) using the formula Row Total x Column Total/Grand Total. I then calculated a chi-square value of 12.78 over a critical value of 3.84 for a degree of freedom of 1 [df=(R-1) x (C-1)].

Total chi-square = 12.78, df = 1
Critical value at .05 is 3.84

A glaring trend obviously exists between the most popular gender of admired leaders by these male and female students. With a 12.78 chi-square value, these trends certainly are NOT due to chance. Only four men in my study admired a leader of the opposite sex the most; whereas 55 women admired a leader of the opposite sex the most. Additionally, 91 total respondents, or 68%, replied that their most admired leader was a male; where as only 43, or 32% of all respondents replied that their most admired leader was a female. Broken down by gender, 10% of men admired a woman most where as 41% of women admired a women most.

The pattern identified may be due to any number of influences. Perhaps because most identified historical heroes are male, both males and females associate leadership with men. In addition, perhaps because males fill most "leadership" positions in society such as ministers, politicians and CEO's, males are associated with leadership. Overall, one may submit that the current patriarchal environment of our society perpetuates an image of men as leaders rather than women. If this is the case, it appears that women (41% admire a female leader most) are breaking the mold of "men as leaders" much more than men (10% admire a female leader most). The Office of Student Life may consider highlighting more past and present women leaders and seek to dissociate a strong leader with a masculine person by advocating for more feminine aspects of a good leader in addition to the regular curricula.

If I were to do this content analysis over, I would attempt to gain a better sample. I would increase my population size (to all of the United States if possible) and ensure that my sample was randomly selected in increase my external validity. I am also skeptical of the disproportionate number of men and women who responded to this survey. If I were to do it again, I would try to ensure that the proportion of men and women who respond are more equal. Perhaps I will do the survey over the phone on through personal interviews rather than through a written survey. After repeating this content analysis a few more times, I would like to develop a research question to examine such as "What is the effect of sexism on the image of leadership?"