Table 2.
Mean composite and subdomain scores for the REFLECT rubric and average general impression scores for 122 medical student essays submitted in 2015 and 2016 via a traditional, private, essay format or social media platform, respectively.
| Private Cohort Scores (95% CI) | Social Media Cohort Scores (95% CI) | t-value (P value)b | |
|---|---|---|---|
| Mean REFLECT Composite Scores (IRR 0.80) | 14.1 (12.0–16.2) | 13.7 (11.4–16.0) | t(1,120)= 0.94, (p=0.35) |
| Mean REFLECT Subdomain Scores | |||
| Writing spectrum (IRR 0.73) | 2.97 (2.5–3.5) | 2.95 (2.4–3.5) | t=0.27 (p=0.79) |
| Presence (IRR 0.80) | 3.12 (2.5–3.8) | 2.86 (2.3–3.4) | t=2.36 (p=0.02) |
| Description of disorienting dilemma (IRR 0.68) | 2.98 (2.5–3.4) | 3.02 (2.5–3.5) | t=−0.44 (p=0.66) |
| Attention to emotion (IRR 0.79) | 2.28 (1.6–2.9) | 2.22 (1.5–2.9) | t=0.50 (p=0.62) |
| Analysis & meaning making (IRR 0.73) | 2.73 (2.3–3.1) | 2.67 (2.1–3.2) | t=0.75 (p=0.45) |
| Mean General Impression Scores (IRR 0.55) | 1.04 (0.7–1.4) | 0.98 (0.6–1.3) | t=1.03 (p=0.31) |
CI, confidence interval; t, t-value; p, p value.
The five Reflection Evaluation for Learners’ Enhanced Competencies Tool (REFLECT) subdomains were scored from 1–4 with a maximum composite score of 20, while general Impressions were scored on a three-point scale (0–2).
a
The inter-rater reliability (IRR) for five faculty raters was calculated using intraclass correlation coefficients.
b
Significance was calculated by comparing private and social media scores using two-tailed t tests.