home
  : Reports : Under-Represented Populations





























home reports instruments plans
search

Under-Represented Populations Annotated Report Excerpts

Return to Under-Represented Populations Reports

Analysis Process

The table below contains report excerpts (right column) accompanied by annotations (left column) identifying how the excerpts represent the Analysis Process Criteria.

Annotations Report Excerpts
 

Excerpt 1 [University of Denver]

Quantitative Analysis:
Describes statistical procedures

Dependent t-tests were conducted on the pretest/posttest differences for each of the 11 questions on the Student Questionnaire of SEM Knowledge and Attitudes. 15 students completed this questionnaire at both time periods, and given the hypothesis that student knowledge would increase in SEM areas, one-tail tests were used (see Table 7 for t-test results).

Presents response rate statistics

Table 7
Dependent t-tests for the Student Questionnaire of SEM Knowledge and Attitudes

TIME GIRLS' COMPLETED SURVEY % PARENTS' COMPLETED SURVEY %
1 18 94.7 19 100.0
2 11 57.9 11 57.9
3 15 78.9 14 73.7
 

Excerpt 2 [University of Washington]

Quantitative Analysis:
Describes procedures for data verification

The completed survey questionnaire forms were entered into a personal computer using R-Base, a standard data-base management package. Preliminary data analyses were used to detect unusual values or patterns and logical inconsistencies. Telephone calls to respondents were made to gather missing answers.

Describes statistical procedures used to organize and reduce the data

Statistical analysis was performed with SPSS/PC+ (Statistical Package for Social Sciences) Version 3.0 on an IBM-compatible personal computer. Data analyses, aimed at describing program characteristics, consisted primarily of summary statistics (means, medians, standard deviations, ranges, and proportions or percents). Two-way tables and correlations were used to describe associations between characteristics. Statistical tests of hypothesis included Chi-Squared, t-tests, and analysis of variance. Confirmation of t-test and ANOVA results was done using non-parametric procedures.

A series of t-tests was used to determine if there were any significant relationships between variables. The t-test was a two sample t-test for unequal variances to test whether or not the means of the two populations were significantly different. Significance was defined at p<.05 to p>-.05. Since a t-test assumes the data come from a normal distribution, the data were also tested using a Mann-Whitney test, which is a non-parametric test. The results of the two tests were similar.

A total of 285 variables were included in six sections of the survey. The six sections corresponded with the six hypothesized underlying constructs/prerequisite conditions for success. Variables within each of the six sections were collapsed into a smaller set of defining variables. At the conclusion, ten to fifteen variables defined each construct/prerequisite condition.

Factor analysis was then executed on each of the six sets of variables, which defined an underlying construct, such as commitment from the Dean. The objective was to detect those variables and/or groups of variables that tended to produce similar responses or response trends and to test the hypothesis that the above mentioned prerequisite conditions or underlying constructs would be present among those institutions with successful intervention programs.

Several linear combinations of variables (factors) were chosen within each construct, because they explained most (>60%) of the variability in the data. These sets of factors were rotated orthogonally to find simpler and more meaningful patterns in which each factor describes the variation shared by a subset of the information highly related to it, and ignores the variation in other, less related variables. Variables were chosen as important components in a factor when correlations between a factor and a variable were greater than 0.5 or less than -0.5.

 

Excerpt 3 [Girls Inc. of Alameda County]

 

Evaluation Results:

The following section presents the results of the Eureka evaluation process. It includes a profile of Eureka participants and data obtained from the participant assessment, focus group meeting, interviews with parents of Eureka graduates and interviews with former Eureka participants.

Quantitative Analysis:
Presents mean response on satisfaction scale

Eureka graduates were asked to rate the overall Eureka Teen Achievement Program on a scale from 1-low to 5-high. Graduates rated the program an overall 4.05.

Qualitative Analysis:
Presents responses to open-ended questions

During the focus group meeting participants relayed examples of ways in which the Eureka program gave them confidence to take on risks or new challenges both in school and in other parts of their lives. A request for specific examples of risks or challenges that they would not have taken had it not been for Eureka elicited these responses:

"I think it (Eureka) gives you confidence, more confidence in going out there and knowing its okay to try your best and that you can succeed."

"They (Eureka) tell you you have to take a risk in life, you know. They tell you … you are a beautiful person and you can only do what you can … It's all about who you are inside and what other people think about you just doesn't matter, and they give you that kind of confidence and you're just like, oh goodness, I can do this or you know if I don't do it this time I can still strive for it next time …"

"They push you to make goals for yourself and set those goals, and then when you reach those goals you make other ones, and you just continue on to new goals."

"Running for office … if you don't have the confidence within yourself that Eureka can help you get, you will not succeed at … High School as Vice President of the Senate … we had an assembly on Friday and I said "Hi everybody, I'm … Vice President of the Senate." I got the loudest yells of anyone, you know what I'm saying? This is the confidence the Eureka has given me."

 

Excerpt 4 [Purdue University]

 

*Graduate Mentoring Program—Engineering: 1994-1997

The Graduate Mentoring Program in Engineering from 1994-1997 has consisted of 190 students. A list of topics, speakers, and an example of formative evaluations utilized throughout the year at monthly meetings for M.S. and Ph.D. women engineering students are included. The one page evaluation sheets contained items that dealt with the number and nature of informal contacts between Mentees and Mentors as well as an assessment of whether monthly events were addressing participants' needs for support, self-esteem, and strategies. Sheets were distributed to attendees at the close of meetings, and findings were used to assess program goals and make program improvements. Table 9 contains group means for each of the monthly meetings held from 1994-1997 as well as overall group means related to specific goals of the program (providing support, heightening self-esteem, and sharing strategies). Responses of participants were on a 5 point Likert scale that ranged from strongly agree (5.0) or agree (4.0) to strongly disagree (1.0).

Quantitative Analysis:
Summarizes survey results in table form

Table 9. Graduate Mentoring Program—Monthly Meetings—Group Means, 1994-1997

Month Support Self-Esteem Strategies
 
94-95 95-96 96-97
94-95 95-96 96-97
94-95 95-96 96-97
August
4.3 4.0 4.0
4.0 4.2 4.0
4.3 4.3 3.9
September
4.4 4.4 4.0
4.4 4.4 4.0
4.4 4.4 4.5
October
4.5 4.0 4.0
4.1 4.1 4.8
4.8 4.6 4.7
November
4.3 4.1 4.3
4.4 4.2 4.3
4.3 4.4 4.4
January
4.6 4.3 3.8
4.5 4.3 4.1
4.6 4.6 4.1
February
4.1 3.9 4.4
4.2 4.1 4.3
4.3 3.4 4.2
March
3.7 4.3 4.3
4.0 4.5 4.0
4.4 4.6 4.3
April
4.4 4.1 4.3
4.3 4.3 4.0
4.4 3.9 4.8
Group Mean
4.1 4.1 4.1
4.3 4.2 4.2
4.4 4.3 4.4

Qualitative Analysis:
Presents findings

Overall group means for monthly meetings, ranging from 4.1 to 4.4, were high. This indicates that program participants agreed that their needs for support, self-esteem, and strategies were being met through the Graduate Mentoring Program in Engineering.

Likewise, a summative survey (see Appendix E) was constructed to examine the background, beliefs, feelings, and future plans of program participants. Results of the 1997 survey are contained in Appendix E. Some of the qualitative and quantitative findings are included below.

Qualitative Responses:
  • "I probably would have quit the Ph.D. program without the peer support from others in the Graduate Mentoring Program."
  • "My [graduate] experience at Purdue has negatively affected my self-esteem. The M&M Program has provided me with positive support and often comments that make me feel O.K. come from associates and friends in the M&M Program."
  • "My interactions with this group have helped me a lot. I appreciate the time and effort that staff members put in to make meetings successful. I think the meetings seem professional and are very productive and of great use to participants."
Quantitative Results:
  • of MS program participants plan to continue on for a Ph.D degree
  • of Ph.D. program participants plan to pursue academic careers

Further, the Graduate Mentoring Program received high scores for providing beneficial meeting topics, using meeting time well, and maintaining relationships between program goals and monthly meeting topics. Participants agreed that the mentoring program was a worthwhile experience for them.

 

Excerpt 5 [University of Washington]

Quantitative Analysis:
Presents results of quantitative data analyses (factor analyses)

Separate factor analyses were conducted on each of the six hypothesized underlying constructs of prerequisite conditions for success. The next section describes the number of factors initially yielded for each of the constructs (eigenvalues >1) and the percent of the total variability explained by each construct.

Construct #1: COMMITMENT FROM THE DEAN

Factor 1     30.3%
  Budget    
  Fundraising assistance    
  Dean evaluates program    
Factor 2     24.4%
  Study center    
  University-level programs    
Factor 3     14.5%
  Training through WEPAN    
  Salary paid by Dean/Provost    
  Dean evaluates program    
    Total variability: 69.2%

Factor analysis on this construct produced three factors, that when taken together, explained 69.2% of the variability. Each factor seems to make intuitive sense. Those institutions with higher budgets tended to have: a) assistance with fundraising; b) evaluations from the deans; c) study centers; d) directors who received training from WEPAN; and e) directors' salaries paid by the dean.

 

Excerpt 6 [Northwest Indian College]

Quantitative Analysis:
Displays quantitative data of primary student retention documents

Student Retention

A simple measure of student retention is the number of students who enroll and complete credits each quarter. In the first quarter, 11 students enrolled, 13 enrolled for the second quarter, and 10 enrolled for the third quarter. One of the students who left between the second and third quarter had severe health problems. When these are addressed, he may return to the program. The other students who failed to complete the third quarter are young males, one of which was a sporadic member of the first TENRM cohort. The retention rate in the first year is 77% based on 10 of the 13 recruits who actually enrolled in the first two quarters.

Student persistence is a measure of student effort. It takes in to account the number of credits students attempt (register to take) compared to the number of credits actually earned during the quarter. Persistence tends to drop as the quarters progress. Figure 3 is a comparison of the mean number of credits attempted compared to the mean number completed. The mean number of attempted credits for all three quarters is 17, where as the mean number of completed credits is 15 for the first quarter, 13 for the second quarter, and nine for the third quarter. Persistence also dropped as the quarters progressed for TENRM I. The completion rate is illustrated in Figure 4. This statistic is important in adjusting for quarterly GPA. Although the rate is just over 50% for the third quarter, some students needed to leave early to begin internships and many students make up work over the summer months.

Figure 3

Figure 4

The five women who entered the program are doing well. All remained through all three quarters and had slightly higher credit completion rates than the males. One of the males who joined the program during second quarter did not complete the third quarter. One left owing to illness. By the third quarter, women comprised slightly less than 50% of the cohort.