Principles Update: Improving our Assessment Accuracy

Overview:

We are excited to share that our personality assessment just got more accurate (thanks to the over 1 million people who have taken the assessment). If you have taken the assessment prior to ~5pm EST on February 7th, 2023, you may now notice small changes to your assessment results. Statistically, these results will be more accurate and therefore individuals who do experience changes should find that their results more accurately reflect their true preferences. Continue reading for more information. 


How we calculate Trait and Facet scores: 

Each trait result that you see is a percentile score, which is your result relative to the distribution of assessment responses from our reference (or “norm”) group. This approach is standard practice in psychological assessments of personality. As is typical in building a research-based assessment, when we first released the assessment, our norm group was based on a sample of approximately 3,000 people.  


What changed:

While this was a statistically significant group to start with, now that we have over 1 million people who have taken the assessment, we had the opportunity to update our norm group to be more representative of the wider population which will lead to more accurate results. Given the size of our current sample, it’s very unlikely we’ll need to make another adjustment to these measures. 


What this means to you:

If you have taken the assessment prior to ~5pm EST on February 7th, 2023 and login to view your results on or after this date, you will see your results calculated relative to this new, larger norm group. This may result in slight changes to you and your team’s assessment results. Statistically, these results will be more accurate and therefore individuals who do experience changes should find that their results more accurately reflect their true preferences. 

Expected impact to the population: 

We expect that some individuals will experience no changes at all, most will experience very minor changes, whereas only very few will experience larger changes. For most individuals, these changes are generally not significant and not likely to change an individual's top 3 archetypes. As such, we expect archetypes to remain the same for most users, but the ordering of archetypes may shift. Some individuals may experience changes up to approximately 25 percentile points for a few traits, however most people will experience much smaller changes for most traits.That said, we do expect that most users should feel their results more accurately represent their true preferences. 

More about percentile scores: 

When looking at your assessment results, each trait score that you see is a percentile score ranging from 0% to 100%, which tells how you answered questions related to that trait relative to the distribution of answers from our updated norm group. 

By way of example, scoring in the 99th percentile means that your answers put you above 99% of individuals within the sampled population for that trait. For instance, if you land in the 99th percentile for Extraversion, this indicates that you exhibit a greater degree of Extraversion compared to 99% of the sampled population. A score in the 50th percentile indicates that your responses place you squarely in the middle of the spectrum on that trait, suggesting that you have an average level of this trait relative to others in the population sample.

You could imagine in the future we may share what are called “local norms” to enable you to compare your results to other more specific populations e.g., college students, or senior executives. This information is not only interesting for research purposes, but could be interesting for individuals as well. We are committed to continuing our analysis of personality traits and continuing to provide what’s most useful to our users to help them most accurately understand themselves and others they relate to.