Closing the Gap in Alzheimer's for African Americans: Analysis of the Forget Me Not Project Questionnaire Open Access

Chapman, Christina Elyse (2016)

Permanent URL:


Older African Americans are about twice as likely to have Alzheimer's disease as older whites. Through the collaborative efforts of the African American Network Against Alzheimer's (AANAA) and playwright Garrett Davis, the Forget Me Not Project works to spread awareness of the growing burden of Alzheimer's in African American communities and to motivate members of minority groups to be more involved in Alzheimer's research and advocacy opportunities. Using the responses to the questionnaire distributed at five Forget Me Not Project events, this study has three main objectives: (1) we seek to summarize how questions in the survey were answered and to summarize the available demographic information; (2) we aim to determine whether questions considered to be indicators of program success are associated with the demographic information in the survey population; (3) we seek to compare methods treating the Likert-scale outcome variables as ordinal, nominal, or binary to determine the best fit for this dataset. To accomplish these objectives, we report descriptive statistics, conduct univariate analyses, and utilize multivariable logistic model selection procedures. The vast majority of program attendees were female and African American. The participants almost unanimously reported that they enjoyed the program, but often responses to other questionnaire items varied by survey site. Within the study population, the extent to which an individual agreed that they learned why older African Americans need to be involved in research was found to be associated with race and the survey location. Also, the extent to which an individual agrees that they would like to learn more about clinical trials was found to be associated with participants' age, survey location, and the participants' willingness to advocate for Alzheimer's research. To find these associations, models treating Likert-scale outcome variables as binary were found to be the most appropriate. In future studies, we suggest implementing changes to the questionnaire, addressing the limitations of the study, and expanding the program to a wider net of communities.

Table of Contents

1 Introduction 1

1.1 Alzheimer's in the United States 1

1.2 Alzheimer's Disproportionate Effect on African Americans 3

1.3 African Americans Network Against Alzheimer's (AANAA) and the Forget Me Not Project 4

1.4 Purpose of This Study 6

2 Methods 8

2.1 Data Collection 8

2.2 Outcomes of Interest 10

2.3 Statistical Methods 11

2.4 Statistical Analysis Software (SAS) 17

3 Results 18

3.1 Results of Descriptive Analysis 18

3.2 Results of Univariate Analysis 20

3.3 Results of Binary Logistic Model Selection 25

3.4 Results of Ordinal Logistic Model Selection 33

3.5 Results of Nominal Logistic Model Selection 38

3.6 Comparison of Three Multivariable Logistic Models 44

4 Discussion 47

4.1 Discussion of Results 47

4.2 Suggestions to Improve Questionnaire 51

4.3 Future Analysis and Limitations 52

4.4 Conclusion 53

5 Appendices 55

A Forget Me Not Questionnaire 55

B SAS Code 56

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Last modified

Primary PDF

Supplemental Files