The format of the interviews was further complicated by the lack of consistent structure in the management of training programs across the NIH ICs. Some have a centralized training officer who coordinates all training (minority targeted and nontargeted); others have a separate minority training officer; still others assign training to individual health scientist administrators depending on the specific research area being funded. In some cases, when training program management was ostensibly centralized, the contractor was told to speak with someone else regarding minority programs. Thus, it appears there is no uniform utilization of minority training programs or training program management policies across the ICs. The inclusion of minority-targeted programs, minority status of participants, career stages served, or whether the program is awarded to an individual or an institution varies considerably across the ICs. For this reason, it was not possible to analyze the response data strictly in terms of career stage. This degree of program management decentralization and heterogeneity made it challenging at times to identify the best respondents for each of the programs. Because some PAIC respondents were responsible for several training programs, it was challenging for the contractor to manage the interviews and ensure that all relevant programs were discussed. Unfortunately, in actual operations, some of the interviews focused on one or more programs that were not even included in the scope of the study.
A total of 22 PAIC interviews were conducted and analyzed for this study. The interviews were administered to a group of rationally selected respondents who administer programs assessed by the study. For reasons of confidentiality, respondent identities and institute affiliations were concealed to anyone other than the contractors who conducted the interviews. The contractor began the interview by introducing the study and explaining how respondents’ comments would be integrated into the study.
The qualitative data analytic program Nu*dist version 6.0 was used to classify response data. Comments were grouped into 347 different clusters of related statements and then subjected to additional analyses by the study committee. The data were organized into four “layers” (or node levels) based on career stage served by the program, whether the program was awarded to individuals or institutions, whether the program was targeted or nontargeted, and the specific topics of discussion.
Common themes are identified and an example quote is provided for each. Rather than duplicating quotations, the prevalence of the response is indicated whenever possible. Unique responses are identified as such. Every effort was taken to ensure accuracy among the quotes reflected in this chapter. Extraneous and irrelevant comments were replaced by ellipses. Words were added to a quotation in cases where missing words could be reasonably discerned from the interview context; added words are included in brackets.