Choosing the best personality assessment
Personality assessments have been popular with training and development professionals, executive coaches, employee assistance program counselors, and career coaches for many years. There is a wealth of assessments to choose from. So what should you look at when selecting the best assessment for your specific need?
It’s unlikely that you will be able to see the survey instrument used by an assessment, but you should be able to discover a few things.
There should be enough questions asked to gather relevant information. A 70-item questionnaire is more reliable than one with only 20 items. Questions should be easily understood and avoid judgmental terms.
Many DISC tests and even DiSC® Classic use a type of forced-choice or ipsative questionnaire. They offer a word or a sentence, and you select which one is most or least like you. These are the least accurate.
Other questionnaires will use a Likert scale, where you rank how well each item describes you on a five-point scale from “never” to “always.” Or, the scale will show choices of “strongly disagree” to “strongly agree” with the statement. This type of question gets more useful answers.
The most accurate instruments will use computer-adapted testing. Questions will vary depending on how a respondent answered previous questions. Wiley now uses adaptive testing for Everything DiSC® after their research showed the following results:
- a 35 percent improvement of reliability scale for people who respond inconsistently
- a 12 percent increase in accuracy over the 79-item assessment (previously used with Everything DiSC profiles)
- a 32 percent increase in accuracy over DiSC Classic
The instrument should result in a report that’s easily understood, memorable, and relevant. Sample reports or profiles are readily available online for most commercial assessments.
Read through a sample. Is it personalized with more than just a name and type or style? Does it look like everyone gets the same report, or is there specific feedback based on the respondent’s results? Does the language of the report present all personalities in a balanced manner? Reports full of charts and graphs can be interesting, but they also need to be memorable and easy to interpret. Is the report something your learners would want to return to in the future?
Are there additional reports available to the administrator or facilitator of the report or profile? If you’re using the assessment for a team, is there a consolidated team report? Are there additional reports available for other applications of the assessment?
Check the reliability and validity properties of the assessment you’re considering. All serious test providers publish them as part of the personality test’s technical documentation. If you cannot find them online or after you contacted the test provider, the test might not be worth purchasing.
Psychometricians have established reliability and validity as the best way to determine if a tool is efficacious. Reliability answers the question: Does the assessment consistently produce the same results across time? Validity answers: Does the assessment measure what it proposes to measure?
An easy thing to check is the assessment’s Cronbach’s Alpha value. It measures the internal consistency of the test, providing an indication of how well the test measures something like an attribute of personality. The rule of thumb says that Cronbach’s Alpha should be at least 0.7 for a test in order to be reliable. As an example, the eight Everything DiSC® scales demonstrate strong reliability, with a median two-week test-retest reliability of .86.
Reliability answers the question: Does the assessment consistently produce the same results across time?
Evidence for validity is typically demonstrated through correlations with other established personality assessments (e.g., NEO-PI-3, 16PF). The validity of the assessment is also demonstrated through correlations with observer ratings of the respondent’s behavior. Documentation for these correlations should be readily available for assessments backed by research. For example, Everything DiSC is meaningfully correlated with two of the Five-Factor Model (FFM) factors: Extraversion and Agreeableness. Despite the similarity in name, “Conscientiousness” in the DiSC model does not have a strong relationship with “Conscientiousness” in the FFM, but rather is measured as a disposition that is analytical and reserved.
Validity answers: Does the assessment measure what it proposes to measure?
How will you use an assessment? Does the assessment address traits that make it relevant to your needs? Will it fit into a learning program you are developing? Are you working with leaders, salespeople, or others who might benefit from a more specialized assessment?
Do you need to complete a certification or accreditation in order to deliver or interpret the assessment? Are there materials available to help you facilitate the profile? Are there facilitators available to hire? Do you require an assessment that will produce a report someone can understand without a facilitator?
You probably get what you pay for, but it’s still important to know that you’re getting the best results within your budget. If you’re providing assessments to more than a few people, look for volume discounts.
Consider costs beyond the assessment, such as printing reports, hiring a facilitator, or purchasing support materials.
Hidden costs for free or inexpensive assessments can include upselling or selling respondents’ email addresses. In other words, some companies will use your contact information to try to sell you additional services or coaching you might not want.
No assessment is able to measure fully every element of a person’s full personality. Assessments vary widely, even those with very similar names. It’s critical to take a careful look before you buy.