---
title: "Prehire Work Experience Predicts Performance at .06. A Practitioner FAQ on Years-of-Experience Requirements."
description: "A 2019 meta-analysis of 81 samples puts the corrected correlation between prehire work experience and job performance at .06, and turnover at .00. Eleven questions on whether a years-of-experience threshold survives scrutiny, what it is legally, and what to measure instead."
canonical: https://www.assessall.com/blog/prehire-work-experience-predicts-performance-at-06-a-practitioner-faq-on-years
published: 2026-11-05
updated: 2026-11-05
author: "AssessAll Editorial"
source: AssessAll
---

# Prehire Work Experience Predicts Performance at .06. A Practitioner FAQ on Years-of-Experience Requirements.

A 2019 meta-analysis of 81 samples puts the corrected correlation between prehire work experience and job performance at .06, and turnover at .00. Eleven questions on whether a years-of-experience threshold survives scrutiny, what it is legally, and what to measure instead.

A years-of-experience requirement is a screening rule that rejects candidates below a stated tenure threshold. Across 44 samples and 11,785 employees, the corrected correlation between prehire work experience and job performance is .06. That makes it one of the weakest predictors in selection, used as one of the hardest filters.

Below are the questions practitioners actually ask when someone proposes removing "5+ years required" from a requisition.

## What does the research actually say about prehire experience?

The most direct evidence is a [meta-analysis by Van Iddekinge, Arnold, Frieder and Roth](https://digitalcommons.unf.edu/unf_faculty_publications/785/) published in *Personnel Psychology* in 2019, which synthesised 81 independent samples. The corrected correlations were:

- **Job performance:** .06 (k = 44, n = 11,785)
- **Training performance:** .11 (k = 21, n = 8,176)
- **Turnover:** .00 (k = 32, n = 11,676)

Two of those deserve a second look. The .00 for turnover means hiring more-experienced people does not, on average, reduce attrition — a common justification for the filter that the data does not support. And when the authors narrowed to measures of *job-, task- or occupation-relevant* experience rather than total years anywhere, the relationship with job performance was still only about .07.

## Didn't earlier research find experience mattered more?

It found something more specific. Quiñones, Ford and Teachout's 1995 review in the same journal reported a positive relationship, but it was strongest when experience was measured as the **amount of actual task experience** rather than the **amount of time in the job**. That distinction is the whole story, and it has survived three decades.

"Five years" measures elapsed calendar time. What a hiring manager means by it is something else: repetitions of the relevant task, exposure to hard cases, and feedback good enough to learn from. Two candidates with identical tenure can differ by an order of magnitude on all three. The clock is not the construct.

## Is experience ever worth weighting?

In three narrow cases:

- **Very early tenure.** Van Iddekinge and colleagues found prehire experience somewhat more predictive of performance when workers have just started a new job — it buys ramp speed, which fades.
- **Task-level specificity.** Task-level experience measures showed some prediction of *training* performance, though on small subsets of studies.
- **When it is scored properly.** McDaniel, Schmidt and Hunter's work on training-and-experience evaluations found the behavioural consistency method — asking candidates to describe past achievements against job-relevant dimensions, then scoring them against anchors — was a considerably stronger predictor than the common "point method" of tallying years and credentials. The point method is the one almost everyone uses.

## Is a years-of-experience filter legally a test?

In the United States, yes. The [Uniform Guidelines on Employee Selection Procedures](https://www.ecfr.gov/current/title-29/subtitle-B/chapter-XIV/part-1607) define a selection procedure at §1607.16(Q) as any measure, combination of measures or procedure used as a basis for an employment decision, and the definition expressly reaches educational and work-experience requirements alongside paper-and-pencil tests. Section 1607.3(A) then states that the use of any selection procedure with adverse impact "will be considered to be discriminatory and inconsistent with these guidelines, unless the procedure has been validated in accordance with these guidelines".

Practically: the threshold on your requisition sits under the same evidentiary burden as a cognitive test, and the four-fifths rule at §1607.4(D) applies to it. Very few organisations have ever computed a selection rate for their experience bar. The EEOC's [guidance on tests and selection procedures](https://www.eeoc.gov/laws/guidance/employment-tests-and-selection-procedures) is the short version.

## Where does the adverse impact come from?

Three routes, none of them exotic:

- **Age, in both directions.** A minimum-years floor screens out younger candidates; a disguised ceiling ("2–4 years, no more") screens out older ones. Both are live [age-discrimination](https://www.eeoc.gov/age-discrimination) exposure.
- **Career interruptions.** Caregiving breaks reduce measured years without reducing capability. The filter cannot tell the difference.
- **Labour-market mobility.** Years accumulated in a market the employer does not recognise often get discounted to zero, which converts a capability screen into a geography screen.

## Are employers already dropping these requirements?

Steadily. Indeed's Hiring Lab [found that in April 2024, 30% of US job postings carried a year-specific requirement](https://hiringlab.indeed.com/2024/05/23/tenure-requirements/), down from about 40% in April 2022. Between April 2023 and April 2024, the share of postings with no years requirement — or an explicit no-experience-needed note — rose from 60% to 70%.

The retreat is uneven. Shares of postings with a year-specific requirement:

**Highest**

- Project management: 49%
- Accounting: 48%
- Civil engineering: 47%

**Lowest**

- Beauty and wellness: 16%
- Pharmacy: 20%
- Therapy: 21%

By education tier, high-education postings fell from 66% in April 2022 to 44%, while lower-education postings moved only from 35% to 29%. The filter is being abandoned fastest exactly where credible alternative evidence is easiest to obtain.

## Does the filter still sort the applicant pool the way it used to?

No, and this is the newest piece of the picture. Hiring Lab's July 2026 analysis found that [in May 2026, 30% of applications to entry-level postings came from workers with 10 or more years of experience](https://hiringlab.indeed.com/2026/07/23/entry-level-jobs-arent-just-for-inexperienced-workers/) — more than from any other experience group. Among those 10-plus-year candidates, 49% of applications went to entry-level roles and only 12% to senior ones. For candidates with 6–9 years, 62% of applications went to entry-level postings.

A years floor was always a weak predictor. It is now also a weak *sorter*: entry-level pools are full of experienced applicants, so the threshold no longer separates the pool into the groups the requisition imagines.

## What about India, where the market is priced in experience bands?

The bands are real, but they move. Naukri's JobSpeak index for September 2026 put overall white-collar hiring up 2% year on year (3,053 against 3,002), with [fresher hiring at +1%, the 4–7 year band at -2%, and demand concentrated above it](https://www.naukri.com/blog/naukri-jobspeak-october-2026-white-collar-hiring-ai-ml-growth/) — 8–12 years at +5%, 13–16 years at +7% and 16-plus at +5%.

That shape matters for screening design. A band that is contracting is full of candidates who will be filtered out of adjacent roles by thresholds set for a different year. Measuring them directly is cheap enough that the filter's cost-saving rationale largely disappears: on AssessAll's pay-as-you-go pricing of ₹30 (about US$0.50) per credit, widening a shortlist by 200 candidates costs roughly ₹6,000.

## What replaces the threshold?

A four-step swap, in this order:

1. **Run a job analysis.** List the tasks the role actually performs, at the level of detail [O*NET](https://www.onetonline.org/) uses. This is the document everything else rests on.
2. **Name the construct the years were standing in for.** Usually it is job knowledge, judgment in ambiguous situations, or a specific technical skill. Write it down. If you cannot name it, the threshold was decoration.
3. **Measure that construct directly.** A job-knowledge test, a work sample, or an AI-graded scenario that puts the candidate in the situation and scores the response against anchors. AssessAll's scored scenarios and integrity-banded proctoring exist for this step.
4. **Keep experience as scored evidence, not as a gate.** Use the behavioural consistency approach: structured achievement questions, rated against dimensions from step 1, contributing points rather than deciding admission.

### If you must keep a numeric threshold

- Write down the job-analysis basis for the specific number
- Make it a preference that adds points, not a bar that rejects
- Set it at the lowest level the analysis supports, not the highest the manager would like
- Compute selection rates by age, sex and any protected group, and test them against the four-fifths rule
- Always pair it with a direct measure, so the threshold is never the only evidence

## How do I know the swap worked?

Run both screens on the same applicants for one cycle: record who the years bar would have rejected, assess them anyway, and compare assessment scores and subsequent performance across the two groups. If the rejected group performs comparably, the threshold was discarding qualified people at no informational gain. The [*Standards for Educational and Psychological Testing*](https://www.apa.org/science/programs/testing/standards) treat that kind of evidence as the basis for validating any intended use of a score — including the use of a cut-off.

## The takeaway

Prehire experience predicts performance at .06, predicts turnover at .00, and no longer even divides the applicant pool cleanly. Replace the threshold with a measure of the thing the years were supposed to indicate, and keep the number only if a job analysis can say why it is that number.

---

Source: https://www.assessall.com/blog/prehire-work-experience-predicts-performance-at-06-a-practitioner-faq-on-years