First-Generation Students in Pursuit of the PhD: Comparing Socialization Experiences and Outcomes to Continuing-Generation Peers

ABSTRACT

Although first-generation students represent a substantial proportion of doctoral students, few studies have examined their experiences and outcomes. We contribute to this nascent area of inquiry by comparing experiences and outcomes of first-generation and continuing-generation students during the first 3 years of doctoral education. Contrary to expectations, the results based on a national sample of PhD students in biology revealed remarkable similarity in experiences and outcomes between first-generation and continuing-generation students. One notable exception to this overall pattern of similarity was research productivity in the second year. By examining the relationships between students’ experiences and outcomes over time, the findings illuminate the unique ways in which socialization experiences are related to specific outcomes and the extent to which those relationships change across years. Thus, this study offers initial insights into the nuanced ways in which students’ socialization experiences contribute to various outcomes of doctoral education.

KEYWORDS:

• Socialization
• doctoral education
• first-generation students
• inequality
• STEM

Notes

1. Students who came from families in which neither parent completed a bachelor’s degree were regarded as first-generation.

2. More specifically, the study focused on “bench biology”—doctoral programs in fields including microbiology, cellular and molecular biology, genetics, and developmental biology.

3. No information was available on first-generation status.

4. The original scale included several additional items, which directly referred to authoring or coauthoring papers for publication and were thus excluded given that one of our outcomes was publication.

5. We combined the three indicators as the overall publication rates were low, especially in the 1st and 2nd years.

6. An alternative specification would involve estimating an HLM model. However, the small number of cases combined with a large number of variables precluded estimating hierarchical linear models (HLM) models. On average, there were only approximately five students per institution in the analytical sample, and many institutions had an even smaller number of students (including a number of institutions with only one or two students). Even baseline HLM models were not able to converge.

7. Attrition was calculated based on the number of students who were no longer in the sample (either because they left the study or left their graduate programs) by the end of a given academic year, which was when the publication survey was administered.

8. We tested interactions between statistically significant doctoral experience indicators and first-generation status for each year and found no statistically significant interactions. Those findings indicate that to the extent that doctoral experiences are related to research productivity, first-generation and continuing-generation students benefit equally.

9. Research infrastructure was also statistically significant at p < .10 in Years 2 and 3.

Additional information

Funding

The work reported in this article was supported by the National Science Foundation through collaborative grants DGE-1431290 and DGE-1431234. The views in this article are those of the authors and do not necessarily represent the views of the funding agency.