Two IQ-style results can differ for more than one reason. Before concluding that your underlying ability changed, check whether you took the same assessment, under comparable conditions, with the same scoring method.

Start with the easiest differences to verify

Write down the test name, version, number of questions, time limit and correct-answer total. If these differ, the two results may not be directly comparable. A shorter test also offers fewer opportunities for a total to distinguish different response patterns.

Check whether both reports show the same kind of value. A domain percentage, an overall score and a percentile are not interchangeable numbers.

Consider familiarity

Seeing similar problems again can change how you approach them. You may recognize an alternating sequence or know how to test a conditional statement. Research on repeated cognitive testing documents retest effects; their size depends on the task and conditions.

This means a higher second score can include improved task familiarity. It does not automatically isolate a broad change in intelligence.

A scoring-version example

Suppose a provider changes its conversion table but leaves your recorded correct answers unchanged. Your displayed score could change solely because the representation changed. That is different from solving more questions correctly.

starp's current test uses a fixed scoring scale. The About page describes the active calibration. When reviewing historical reports, distinguish changes to the scoring method from changes in your performance.

Review the session itself

Were you interrupted? Did the timer expire? Did you misunderstand a rule or accidentally select an option? Specific observations are more useful than guessing a single cause from the final number.

Keep a small practice log and look for repeated patterns over several comparable sessions. If numerical errors consistently involve percentages, work on that skill. A score difference becomes more informative when you connect it to what actually happened during the questions.