Mentorship Findings: Qualitative Analysis of Mentorship

According to the PLOS One study PLOS One by Loyola Irizarry et al., 2026, most undergraduate biology and biomedical science students reported having at least one mentor, but the sources and patterns of mentorship differed by institution.

• In Spring 2024, Loyola Irizarry et al. surveyed 1, 327 students at Institution A and 1, 174 students at Institution B, providing the dataset for this analysis.
• In August 2026, the PLOS One paper reported that 55% of students at Institution A and 65% at Institution B said they had a mentor, a difference significant at p = 6.92 × 10−8.
• The PLOS One study found that mentors outside the institution were the most common source (38.1% at Institution A, 43.1% at Institution B) and peer mentors were reported by 25.2% at Institution A versus 34.4% at Institution B.
• Loyola Irizarry et al., 2026, report that perceived mentoring support clustered into two validated domains: Degree and Career Support and Existence of a Role Model, with mean scores near 5.0 on a 6-point scale.

The PLOS One study directly answered where undergraduate BBS students obtain mentorship and how they perceive support by administering a mentoring-focused questionnaire in Spring 2024, according to Loyola Irizarry et al., 2026.

Loyola Irizarry et al. asked students to nominate the mentor who most influenced their career development and then measured experiential similarity, interaction frequency, relationship length, and perceived support using adapted items from the College Student Mentoring Scale (CSMS).

The study used confirmatory factor analysis and measurement invariance testing to validate two CSMS subscales across both institutions, finding configural and metric invariance but not scalar invariance, so correlations are interpretable across groups but mean comparisons require caution, per Loyola Irizarry et al., 2026.

For qualitative researchers, the PLOS One dataset shows that mentorship networks are multi-sourced and that single-mentor surveys can mask complexity, according to Loyola Irizarry et al., 2026.

• When designing instruments, include prompts for multiple mentors and context tags (in‑institution, outside, peer, formal program), because Loyola Irizarry et al. found students often report mentors outside their institution.
• When analyzing open-text mentor descriptions, code for mentor role, formality, and career alignment: Loyola Irizarry et al. measured experiential similarity and found nearly half of students reported sharing career field with their mentor.
• Because measurement invariance failed at the scalar level in the PLOS One study, triangulate survey scores with qualitative themes before making cross-institution mean comparisons.

AI-enabled qualitative analysis makes mentorship networks visible at scale and complements the survey-based findings reported by Loyola Irizarry et al. in PLOS One (published August 19, 2026).

The PLOS One study's Spring 2024 data (n = 1, 327 at Institution A; n = 1, 174 at Institution B) shows clear opportunities for program evaluation, peer‑mentor scaling, and mapping external mentor pipelines.

If you want to replicate thematic coding, cross-segment comparisons, and extract quotable evidence from mentorship surveys, try an AI-first pipeline built for qualitative research.

Get started and Try Evidano for free to upload transcripts, run thematic and cross-segment analyses, and produce extractable quotes and visualizations for reports.