If you take a GRE practice test and want to know what your raw score means for the official GRE scaled score, you need a conversion process. This article explains how GRE practice test score conversion works, why conversions differ across test providers, and gives a step-by-step method you can use to convert raw practice results into realistic scaled-score estimates that help guide prep decisions.
Quick answer
Raw-to-scaled conversion depends on test difficulty and the scoring model used by the test maker. The most reliable conversions come from official ETS practice tests (PowerPrep). For third-party practice tests, convert raw scores to percentiles, then map percentiles to ETS scaled scores, or build an empirical conversion table by taking multiple tests and averaging results. Use conversion as an estimate, not an exact prediction.
Why this matters
Admissions decisions and target scores are stated in official scaled scores (130 to 170 for Verbal and Quant). Knowing how a practice test raw score maps to that scale helps you:
- Assess whether you are on track for your target score.
- Decide which areas need focused improvement.
- Compare performance across different practice resources.
- Set realistic study goals and timelines.
Common challenges students face
- Different providers use different scales. ETS’s PowerPrep is calibrated to the real GRE. Other companies use their own equating and may overestimate or underestimate scaled scores.
- Small sample sizes. A single practice test can be an outlier. Difficulty can vary by test edition, producing different raw-to-scaled relationships.
- Ignoring adaptive structure. The GRE’s section-level adaptive format means performance on the first section can influence question difficulty on the second, which affects conversions.
- Mistaking raw-correct counts for scaled equivalence. The number of correct answers needed for a given scaled score depends on the specific test form.
Essential GRE scoring facts (short)
- Sections scored 130–170: Verbal Reasoning and Quantitative Reasoning are each reported in this range with 1-point increments.
- Raw score basis: For each of Verbal and Quant, your raw score equals the number of questions you answered correctly across both scored sections (typically 40 questions total per measure, two sections of about 20 items each).
- Analytical Writing: Scored separately 0–6 in half-point increments; conversion methods below focus on Verbal and Quant.
Step-by-step strategy to convert practice-test raw scores to GRE scaled scores
-
Start with ETS PowerPrep when possible.
PowerPrep practice tests are produced by the same organization that makes the real GRE. If you take PowerPrep, treat its scaled scores as the baseline. Use those tests to see how raw correct maps to ETS scaled scores for you under test conditions.
-
Record detailed raw data consistently.
For each practice test, log: raw correct in Verbal and Quant, time management notes, number of educated guesses, and whether sections felt easier or harder than previous tests. Consistent logging lets you spot trends and outliers.
-
Convert third-party test raw counts to percentiles.
Many third-party services report percentiles or provide their own scaled score. If they report percentiles, you can map those percentiles to ETS scaled scores using the ETS percentile-to-scale relationship (see the method in step 4). If percentiles are not given, you must create an empirical mapping (step 5).
-
Map percentiles to ETS scaled scores.
ETS publishes percentile charts that show how scaled scores correspond to percentiles. If a practice test gives you a percentile for a section, find the ETS scaled score with the closest percentile. This mapping gives a practical estimate without relying on raw-count conversions that vary by test form.
-
Create an empirical raw-to-scaled conversion table for non-ETS tests.
Use this process:
- Take several full-length tests from the same provider under realistic conditions.
- For each test, note your raw correct and the provider’s reported scaled score or percentile.
- For each raw-correct value, average the provider’s scaled scores across multiple test forms to smooth noise.
- Compare the provider’s averaged scaled scores to ETS percentiles or to a PowerPrep baseline. Adjust the table if you see systematic bias (for example, the provider consistently reports scores 1–3 points higher than PowerPrep for the same raw counts).
After you build at least one table, test its predictive validity by taking a different test from the same provider and seeing how close your predicted ETS-scale matches any ETS-based benchmark you can get (PowerPrep or percentiles).
-
Use smoothing and ranges, not precise single-point estimates.
Because forms vary, convert raw counts to a scaled-score range (for example, 158–160) rather than a single number. This acknowledges uncertainty and prevents overconfidence.
-
Update conversions as you accumulate data.
Every new test result refines the mapping. After 4–6 tests from a provider you will have a more dependable conversion than after one test.
Hypothetical example: converting a third-party practice raw score
The following example shows the method. This is illustrative and not official ETS data.
| Step | Example |
|---|---|
| Take two practice tests from Provider X | Test A: Verbal raw correct = 28 of 40. Provider X reports scaled 162. Test B: Verbal raw = 26, scaled 160. |
| Average scaled by raw | Raw 28 → 162; Raw 26 → 160. If you later have Raw 27 → average 161. |
| Compare to ETS percentile mapping | Provider’s 162 corresponds to roughly the 93rd percentile in their system. Find ETS scaled score with similar percentile. If ETS 162 ≈ 92nd percentile, good alignment. If ETS 162 ≈ 95th percentile, adjust by 1–2 points. |
| Predict for a new test | If on a new Provider X test you score raw 27, predict ETS-scale about 161, but give a range 160–162 to allow for variation. |
Practical tips for trustworthy conversions
- Use ETS PowerPrep as your anchor. If you only use one provider, make it ETS when possible. PowerPrep scaled scores are the most realistic proxy for the actual GRE.
- Prefer percentiles where available. Percentiles are more stable across test forms than raw-to-scale relationships. Mapping percentiles to ETS scaled scores reduces error.
- Take multiple tests from each provider. One test is noisy. Two or more builds confidence. Use averages and ranges.
- Match practice difficulty to target test conditions. A test that consistently feels easier may inflate your conversion. Simulate real test timing, breaks, and conditions.
- Track item-level mistakes. Raw counts are useful, but item-type patterns reveal whether errors are careless, conceptual, or timing-related. A raw-to-scale number alone does not diagnose weaknesses.
- Recalibrate after major improvements. If your study drastically improves performance style (for example, reducing careless errors by changing timing strategies), your prior conversion table may need updating.
Common mistakes to avoid
- Assuming one provider’s scaled score equals ETS scaled score. Third-party providers do not necessarily use the same equating as ETS.
- Relying on a single practice test. You risk over- or underestimating your true ability.
- Using raw counts without considering difficulty. Two tests with the same raw correct can lead to different scaled scores if one included more difficult questions.
- Ignoring variability across test forms. Treat conversion as probabilistic, not deterministic.
How to implement this in your study plan (a 6-week example)
-
Week 1 — Baseline
Take an ETS PowerPrep test under timed conditions. Record raw correct for Verbal and Quant and the ETS scaled scores. This establishes your anchor.
-
Weeks 2–3 — Practice and build conversion tables
Take two different third-party full-length tests (one per week). For each test, log raw counts and the provider’s scaled or percentile results. Start a simple table mapping raw counts to provider scaled scores.
-
Week 4 — Cross-check
Take another PowerPrep test. Compare your predicted ETS scores from provider tests with this PowerPrep result. Adjust your conversion table for systematic bias.
-
Weeks 5–6 — Final tuning
Use your refined conversion to set final study targets. Focus on the item types that most often lower your raw counts. Continue taking one full practice test per week, and update conversions.
Sample conversion table (for illustration only)
This table is a hypothetical example showing how raw correct might map to scaled scores after smoothing and averaging. Do not treat these numbers as official. Build your own table using the steps above.
| Raw correct (of 40) | Estimated ETS scaled score | Suggested confidence range |
|---|---|---|
| 36–40 | 169–170 | 169–170 |
| 33–35 | 165–168 | 164–169 |
| 30–32 | 161–164 | 160–165 |
| 26–29 | 156–160 | 154–162 |
| 22–25 | 150–155 | 147–157 |
| 18–21 | 144–149 | 141–150 |
| 0–17 | 130–143 | 130–145 |
Frequently asked questions
Q: Which practice tests give the most accurate scaled-score estimates?
A: ETS PowerPrep is the most accurate because it is produced by the test maker. Among third-party providers, accuracy varies. Look for providers that publish percentile data and that have been independently evaluated. Always anchor your conversions to an ETS PowerPrep result when possible.
Q: Can I convert raw scores from a paper test the same way?
A: Yes. The conversion principles apply whether test is paper or computer. For the current computer-delivered GRE, Verbal and Quant are reported on the 130–170 scale, so map raw correct totals the same way. If your paper test is not designed to mirror GRE content and difficulty, its raw-to-scaled mapping will be less reliable.
Q: How many practice tests do I need to create a reliable conversion?
A: Aim for at least three tests per provider, more if possible. Two tests are a minimum to spot big biases. Four to six tests give a much stronger basis for smoothing and producing reliable ranges.
Q: Why do some practice providers report higher scaled scores than PowerPrep?
A: Providers may create item pools that are easier on average or use equating methods that yield higher scaled scores. They may also calibrate to older data or use different anchoring. That is why direct comparisons without calibration are risky.
Q: Should I obsess over a 1-point difference in predicted scaled score?
A: No. Treat conversion estimates as ranges. A one-point variance is within normal measurement error. Focus on consistent improvement in raw correctness and on reducing frequent mistake types.
Final thoughts
Converting GRE practice-test raw scores to official scaled scores is a practical skill, not an exact science. The most reliable approach is to anchor conversions to ETS PowerPrep, use percentiles when available, and build empirical mappings from multiple tests per provider. Always report results as ranges, update your mapping as you get more data, and use conversions to guide focused study rather than as definitive predictors. With careful logging and repeated testing, your conversions will become a useful tool for tracking real progress toward your target GRE scores.
