Structured reference checks predict job performance. Here's the research.
Reference checks earned a bad reputation from letters and unstructured phone calls. Decades of research show that structured, job-related, multi-source references measure something real about how people perform at work.
What the studies found
Correlations (r) between pre-hire reference ratings and later outcomes. No single hiring method predicts perfectly; these numbers show meaningful signal, not certainty.
Web-based references
Correlation between multi-source reference ratings and later supervisor ratings (Hedricks, Robie & Oswald, 2013).
Structured phone references
Corrected validity against supervisor ratings for 223 entry-level hires (Taylor et al., 2004).
Combined with ability tests
Validity of cognitive ability plus reference checks, up from .51 for ability alone (U.S. OPM).
Reliability
Internal consistency of structured web reference ratings across about 8,000 cases.
Two very different things are both called "references"
A 1979 review found traditional reference reports unreliable and lenient. That criticism was about the method, not about references themselves.
| Traditional reference | Structured reference |
|---|---|
| Free-form conversation or letter | The same job-related questions for every candidate |
| Topics vary by recruiter and referee | Defined competencies and behavior-focused questions |
| Narrative impressions | Rating scales plus targeted comments |
| Often a single reference | Several references combined |
| No consistent scoring | Standard scales and report format |
| Manual calls and notes | Digital collection, reminders and timestamps |
A real signal, best used alongside other evidence
- They predict performance. Structured reference ratings relate to later supervisor ratings, and higher scores go with lower involuntary turnover.
- They add information. References give a third-party view that interviews and tests don't fully capture.
- Structure reduces bias. A 2022 study of more than 4 million structured reference ratings found negligible practical gender bias.
- They are low cost. Digital collection makes the same process fast and repeatable for every hire.
Leniency is real, and it can be reduced
A 2026 series of eight experiments with more than 1,700 managers found they rate former employees more favorably in reference checks than in private. Asking referees to answer from the hiring manager's point of view reduced that leniency.
References support a hiring decision. They don't make it, and no reference check can guarantee a hire will succeed.
How OnlineReferenceCheck.com applies it
Job-related questions
Role templates focus every question on observed work behavior, and every candidate gets the same set.
Multiple observers
Up to 6 references per applicant, so no single rater dominates the result.
Scales plus comments
Rating scales make candidates comparable; comments show the reasons behind the numbers.
Leniency guidance
Referees are asked to answer as if they were the hiring manager.
Traceable summaries
Every AI summary links back to the original answers, and agreement between references is shown.
Human decisions
Reports and fraud flags are review prompts for your team, never automatic hire or reject decisions.
The Scientific Case for Structured Reference Checking
An 8-page review of the research on reference checking: predictive validity, reliability, fairness, the leniency problem, and what makes a reference check defensible.
- Findings from 10 peer-reviewed studies and U.S. OPM guidance
- What the research supports, and the claims it doesn't
- A 9-point checklist for a scientifically defensible reference check
Research cited
- Hedricks, Robie & Oswald (2013). Web-based multisource reference checking. International Journal of Selection and Assessment, 21(1).
- Taylor, Pajo, Cheung & Stringfield (2004). Dimensionality and validity of a structured telephone reference check procedure. Personnel Psychology, 57(3).
- Fisher et al. (2022). Little cause for concern: Gender effects in structured employment references. International Journal of Selection and Assessment.
- Jacobson et al. (2026). Vicarious self-promotion underlies managers' leniency in online reference checks. International Journal of Selection and Assessment.
- Schmidt & Hunter (1998). The validity and utility of selection methods in personnel psychology. Psychological Bulletin, 124(2).
- Muchinsky (1979). The use of reference reports in personnel selection. Journal of Occupational Psychology, 52.
- U.S. Office of Personnel Management. Designing an Assessment Strategy; Reference Checking.
These findings come from published research, not a validation study of OnlineReferenceCheck.com.
Put structured references to work.
See a structured reference check from request to scored report in a 30-minute demo.