How To Create A Personality Test & Cultural Fit Assessment

A personality test is a method of assessing human personality constructs (aka. personality trait dimensions, facets, factors, or aspects).

Personality assessment instruments are introspective self-report questionnaires which constitute of rating scales (a series of multiple choice questions). That is why personality inventories are classified as Subjective Tests. The output depends on the examination of one's own conscious thoughts and feelings - which is a subject’s self evaluation.

The first personality assessment measures were developed in the 1920s and were intended to ease the process of personnel selection (aka. Hiring, Recruitment), particularly in the armed forces. Since these early efforts, a wide variety of personality scales and questionnaires have been developed.

Today questionnaires have gained huge popularity across the globe in the area of personality assessment because of its ease of construction and administration, despite its serious limitations such as social desirable responses, judgmental errors, lack of adequate self-insight (or biased perceptions of others) and faking etc. (Faking is defined as deliberate provision of “inaccurate responses to personality items in a manner that they believe will increase their chances of obtaining valued outcomes, such as a favorable hiring decision”).

Estimates of how much the personality assessment industry in the US is worth range anywhere from $2 and $4 billion a year (as of 2013). Personality assessment is used in a wide range of contexts, including individual and relationship counseling, clinical psychology, forensic psychology, school psychology, career counseling, employment testing, occupational health and safety and customer relationship management.

A survey of 84 large organizations (with an average annual revenue of 45.7 billion dollars) found that

• 66% of the organizations used personality testing for “high potential” employees
• and 57% used personality testing for senior executives.

Personality testing is now the second most common pre-employment assessment practice. Further, 74% of US-based HR professionals believe that personality testing can be useful.

Most of the instruments used today such as Minnesota Multiphasic Personality Inventory (MMPI), The Sixteen Personality Factor Questionnaire (16PF), The Comrey Personality Scales, and Revised NEO Personality Inventory (NEO PIR) are based on the Big Five and Five Factor Model of personality.

1. Scoring of Personality Tests

A personality test is a self-report inventory which requires respondents to introspectively assess their own personality characteristics by indicating their level of agreement with each item. An item on a personality test, for example, might ask respondents to rate the degree to which they agree with the statement "Don't talk a lot" on a scale from 1 ("strongly disagree") to 5 ("strongly agree")

Unlike other kind of assessments (such as aptitude tests or language tests) questions in Personality Tests do not have a correct / incorrect answer. However;

• Each option in the scale is mapped to a Trait (or a series of Traits) by a different scoring. For instance the option "Strongly Agree" may be assigned to the Extraversion Trait by a score of 1 (or 5 if the keys are reversed).
• Traits may be structured (aka. grouped by) in a tree-like form so that when an option that has been assigned to a trait has been selected by the respondent; both the trait and the parent trait(s) should be affected by the score. For instance in the IPIP NEO PIR Test, the traits Activity Level, Assertiveness, Cheerfulness, Excitement-Seeking, Friendliness and Gregariousness are grouped by the Extraversion Trait.

So when the respondent answers a question, the scores of the traits that are assigned to the selected option are affected. When the respondent answers all the questions in the personality test, all traits will have their own scores calculated.

2. Using Norm Groups to Derive Results

Even thought personality tests are scored by scoring respondent's answers for each trait, the scores for each trait do not say much about the respondent personality by itself. (Personality test scores are difficult to interpret in a direct sense). That's why the scores for each trait must be compared with the rest of the population who have already completed the personality test. This is where the norm group comes into play. Norms provide a comparative basis for interpreting a respondent's test scores. Common formats for these norms include percentile ranks, z scores, sten scores, and other forms of standardised scores. By using a norm group, the test administrator can gage how your traits (for example Extraversion) compares to the rest of the population.

To interpret individuals' scores, one might calculate the mean and standard deviation (SD) for a sample of persons, usually of the same sex and a particular age range, and interpret

• scores within one-half SD of the mean as "average."
• scores outside that range can be interpreted as "low" or "high."

If the scores are normally distributed, this would result in approximately

• 38% of persons being classified as average
• about 31% as low
• and 31% as high

An alternative method for showing respondents where they stand with respect to a group of respondents is to divide a set of scores into five equal parts, which are called quintiles. Labels for the scale anchors describe the lowest 20% and highest 20%, the label "average" is used for the middle 20%, and the remaining quintiles are labeled "somewhat." For example, labels for the quintiles on a scale that ranges from introversion to extraversion would be introverted, somewhat introverted, average, somewhat extraverted, and extraverted.

3. Profile Matching & Benchmarks

Norm Groups are used to derive meaningful results given a set of trait scores. However when using personality tests or cultural fit assessments for any kind of selection process (employee selection, school selection, career selection etc..), using Norm Groups are not enough. Because it is not enough to know about the personality of the respondent, but also will you need to find out how much the personality fits into the position/job or organization's culture.

In order to derive such conclusions, benchmarks - for examples job specific benchmarks - that assess how good a fit a person's personality is for a given position should be created. The benchmarks calculate a person's job match score by comparing their personality trait scores to composite profiles created

4. Score Checks & Correction for Faking

To derive dependable results, there will usually be some kind of ‘score check’ involved, so that the test administrator can see things such as how consistently the respondent answered the questions or whether he/she just went for the neutral option on most of them.

If you want to create and deliver personality tests online, you will need an online assessment software that is capable of

• Creating a Trait Hierarchy
• Creating Multiple Choice Questions having each option assigned (aka. mapped) to one or multiple traits.
• Creating custom reports for each test
• Creating custom matching algorithms (aka. functions) for benchmarking that returns match scores.
• Allowing you to create a customizable algorithm (aka. function) for each test, to handle the correction for faking.

1. Add Effects & Map Traits to Questions

The Multiple Choice Question Editor allows you to map each option to one or many traits (aka. effects). That allows to track & report automatically each option.

For example in the screenshot below, you will observe that the options of the questions are mapped to Agreeableness/Politeness trait with a different scoring each.

• If Very Inaccurate is selected, Agreeableness/Politeness will be increased by 1
• If Moderately Inaccurate is selected, Agreeableness/Politeness will be increased by 2
• If Neither Accurate nor Inaccurate is selected, Agreeableness/Politeness will be increased by 3
• If Moderately Accurate is selected, Agreeableness/Politeness will be increased by 4
• If Very Accurate is selected, Agreeableness/Politeness will be increased by 5

2. Create Custom Reporting Template

Each personality test is different. Hence the reporting. That's why Testinvite allows you to have your own reporting template. The template is a simple HTML template with placeholders.

3. Create Norm Groups & Benchmarks & Match Scoring Algorithm

You can create your own function (algorithm) to create norm groups, benchmarks and match scores.