The research supplement will focus on social climate, as expressed through message linguistics, subjective experiences, and social networks. The top-down assessment will use discourse analysis to explore unfiltered e-mail communication regarding women in STEM. The connection between the top-down and bottom-up processes will be assessed by analyzing the depth and value of STEM professors’ networks and administrative expectations of women’s networks. These studies will be conducted by faculty and graduate students, and will be turned into a series of presentations and publications.
Study 1: Analyzing top-down communication through discourse analysis
It is increasingly acknowledged that discourse is the basis of the organizational environment (e.g., Alvesson & Karreman, 2000; Taylor & Cooren, 1997), with these discourses embedded with cultural expression (Alvesson & Karreman, 2000; Fairhurst, 2004). The structure of communication determines how an issue is discussed, behavior during conversation, and resulting knowledge (Hall, 2001). In turn, communication can result in larger schemas about self and others (Fairclough and Wodak, 1997) becoming instrumental in directing related behaviors. Language can therefore be studied to reveal the underlying power structures and ideology of an institution and the pathways through which it influences institutional outcomes (Hardy, 2004).
Discourse analysis examines how language is used to communicate knowledge, identity, and social relationships within an institutional framework (Luke, 1996). This technique is commonly used to investigate how power relationships are transmitted and social identities are constructed (Van Dijk, 1993). Discourse analysis avoids the pitfalls of self-report by analyzing real world communications. When using messaging through electronic mediums like e-mail, people are less likely to consciously control for subtle biases and dominance may be manifested through verbal mechanisms such as rhetoric, semantics, turn-taking, formality, and positive and negative representations (Van Dijk, 1993). UC LEAF can gain understanding of the underlying beliefs of the organization and the information transmitted to its members by examining commonplace interactions.
A number of electronic documents will be sampled, extending from two years prior to UC LEAF, through the first four years of the project. All STEM-related online communications (e.g., e-mails and webpages) will be examined for relevance. All communications to STEM scientists via email distribution lists will be examined, as well be workload documents, departmental and college bylaws, and reappointment, promotion and tenure policies. Throughout the first four years of UC LEAF, this corpus will be continuously updated. Data coding will be developed through informed adaptation of the grounded theory approach that involves using inductive reasoning to generate theory of an organization originally advocated by Glaser and Strauss (1967). More specifically, the current project will employ a grounded theory perspective later modified by Strauss that allows the researcher to start with basic research questions and problems (Corbin & Strauss, 1990; Devadas, Silong, & Ismail, 2011). This perspective has a constructivist approach which encourages the researcher to take into account context and the subjective experiences of the participants, something typically restricted in the Glaserian tradition (Corbin & Strauss, 2008; Devadas, Silong, & Ismail, 2011). Theory is based on data, not a priori hypotheses. By adapting theory to identifiable real world occurrences, it is made more functional within the context of a given organization. We will therefore identify patterns of communication to gain a clearer picture of social interactions within UC, and base theories of change off of this information. In addition to examining the experiences of women, data analyses will focus on messages which deal directly and indirectly with racial and ethnic minorities, and grounded theory will be carefully applied to these various subgroups.
A baseline of communication will be assessed by examining data from the two years previous to NSF ADVANCE funding. It is hypothesized that following this period, social inequality, as expressed through written language, will decrease. Conversely, supportive communications toward women in STEM will increase. Satisfaction with work, based on annual surveys from women in STEM, will be expressed in regular communication, correlating with unit level differences in communications. These changes in communication will vary by STEM unit in ways that are explicable from other data (logic models, key informant interviews, etc.).
Study 2: Analysis of Network Depth and Value
Social relationships and intra-organizational networking have repeatedly been shown to predict career success (e.g., Sparrowe, Liden, Wayne, & Kraimer, 2001). However, building social networks may become difficult when STEM women enter into an area where they are judged, implicitly or explicitly, to be less competent outsiders. In addition, social networks become more difficult to manage when work-life balance becomes unmanageable. Thus, a key component to analyzing the success of UC LEAF entails assessing the social climate through real connections between people within UC.
All women involved in learning communities during years 1-4 of the grant will be asked participate in outlining their networks annually. At the end of the study, we will have four years of data from LC participants in year one, three years of data from participants in year two etc. To compare men and women’s networking success, we will invite men to participate. This subgroup of men will be nominated by women peers in their department. If they decline, we will work with the women participants to find another match.
Once a year, each participant will be invited to send us a list of potential references. To be included on a reference list, participants must be at least 90% certain that the nominated references would write a letter of recommendation for him or her. We will prompt the women to include names of references who could write in response to narrow prompts such as “community leadership,” “university service,” etc. in addition to names of people who could provide comprehensive career progress and impact evaluations.
Beyond reporting potential references, participants will submit a list of support people. Participants may submit as many names as they wish, but each person listed should be someone that the person knows he or she could rely on for help or support. In addition to a general category of professional career support, we will prompt for specific kinds of support people, such as “proposal editing,” “conflict with supervisor,” and “advice on work-life balance.” When a reference or support person is not a UC employee, the participant will be asked for their title and organization.
After writing their lists, participants will be asked to complete a short questionnaire about their degree of satisfaction with their reference and support groups. Participants will also be asked about the utility of their networks in achieving life and career goals. They will be prompted not to evaluate specific individuals, but rather to rate their degree of satisfaction with categories of references and support people and the usefulness of those categories for goal attainment in specific domains such as publishing, funding, career advancement, etc. In addition, participants will be asked each year to indicate what three steps they plan to take in the coming year to increase the depth and value of their network. The following years, we will feedback these answers to the participants on yearly basis (but only these answers, not names generated) and ask the women to evaluate the success of their planned efforts.
Administrators will be asked to make predictions about reported network depth and value among men and women STEM scientists. Approximately 10 department heads, directors, and deans will be interviewed. Questions will inquire about their perceptions about women’s satisfaction with work, amount of supportive communication, and network equality between the genders.
While it is hypothesized that men in year one will have more valuable social networks, it is anticipated that the work of UC LEAF will over time lead women to have networks that closely resemble their male peers. In year one, men will likely have richer, deeper, and self-assessed more valuable support networks, as well as support networks that include a larger number of more senior colleagues. However, women participants in UC LEAF programing will close the above gaps with their male peers over the course of the research program. Overall, STEM women’s supportive communications will increase over study period, and this increase will correspond with women scientist satisfaction with work based on annual surveys. Change in communications will vary by STEM unit in ways that are explicable from other data (logic models, key informant interviews, etc.)
As with the first two studies, racial and ethnic minority women will be assessed as subgroups as well as with other STEM scientist women. It is predicted that minority women will initially have greater discrepancies in network depth and value. Over time, these networks will also become more similar to those of other women and men. Regarding the administrators interviewed, it is anticipated that they will overestimate equality in networks between the genders, the utility of women’s networks, and amount of supportive communications. However, it is believed that women’s networks will become increasingly in line with administrative expectations.