Psychometric Assessments are used extensively by HR and L&OD professionals, managers and leaders for making hiring decisions and creating individual developmental plans. So far, its benefits have always been recognized by stakeholders who have used it.
However, there is still great confusion to identify and determine which psychometric assessments to use.
Typically, when selecting a psychometric assessment tool, ask the following questions:
- What is the purpose and objective?
- Why am I looking for a psychometric assessment tool?
- How do I plan to use and when?
- Who is my target audience?
- What does the tool assess?
- What are its psychometric properties – in terms of reliability, validity and norms?
- How much time can the participants spend in taking the assessment?
- What is the outcome (in terms of the report) am I/are we expecting?
- What is my budget?
If you can find answers to these questions, chances are that you will find the best fit psychometric assessment for yourself.
Most of the questions listed above are simple to understand. Question 4, however, needs more explanation and awareness so that one finds the right information to make decision.
The reliability and validity and norms are the internal attributes of an assessment and are collectively called Psychometric Properties. Here are few more details about it:
Reliability: It refers to “consistency”- how consistent is the assessment tool and what is the degree to which the assessment is free of measurement error. The reliability value is assessed by a statistical computation method named Correlation and when looking for the reliability of the assessment, look for the “r” value. This value cannot be greater than plus minus 1. A reliability score of 0.70 and above is considered is good reliability and higher the better.
There are various methods of assessing reliability, but the two most used ones are Test-Retest Reliability and Internal Consistency Reliability. The test-retest reliability assesses the consistency over a period (say, a fortnight) to assess if the assessment gives consistent results under repeated measurement conditions. Internal consistency, on the other hand, is the consistency within the items/questions and the subsequent competency buckets it is assessing.
Validity: It gives an indication of whether the assessment measures the construct what it claims to measure. The most common type of validity is face validity – basically, if you read the items/statements, you should be able to indicate what it is aiming to assess. This is the most basic type of validity and does not require any complex statistical evaluation.
Apart from that, the other three validity types evaluated for an assessment are
- Content Validity: It measures and indicates whether the items in the assessment are representative of the domain/construct it is assessing. For example, if the assessment is for emotional intelligence (EI), it should not just assess the intrapersonal aspect of EI but also the interpersonal aspect. When both these dimensions are assessed by the assessed, it is representative of the construct (in this case, EI) and the content validity will be high.
- Construct Validity: It indicates if the scores produced from the assessment is aligned to how the tool should behave. For instance, if you planning to assess emotional intelligence (EQ), then the score should indicate the degree to which a person is emotionally intelligent.
- Criterion Validity: It evaluates if the assessment is an effective predictor of whatever it is trying to assess. For example, if the EQ score is low then the person may get too stressed under pressure, may not handle emotionally laden situations effectively etc. Or, for instance, if a person has a high IQ, then they can effectively cope with the environment, process information faster and reason out well.
Norms: A score on the assessment will be meaningless if it can’t be interpreted using a reference point. What would a score of 56 on EQ assessment or a score of 78 on aptitude test mean anyway? But, if we say that high emotional intelligence is indicated by scores 70 and above or that a score of 78 falls in the bracket of 75-100 and it means high aptitude, then it makes much more sense. This reference point is called norms. The norms are derived from a sample group of individuals who are representative of the target group for whom the assessment has been created. Within that, we can create norms based on age, geographical location, profession etc.
Knowing the target audience for your purpose (as mentioned in the questions above) is important to ensure that the assessment’s norm also has been standardized on that target audience. If your target audience are Indian in the age group of 20 – 35 age group, you need to check if the norms of the assessment are also aligned similarly. If not, the score and its interpretation will not be accurate. You can’t assess the score of a 55-year-old with a norm group score of 20 – 35 years age group because emotional intelligence and aptitude would mean different and differ in degree for the two groups.
Once you have checked for these three key psychometric properties, you can be rest assured that you have chosen a good tool. If you or the candidates then question what the report of the assessment indicates, then you are questioning the psychometric properties of the tool. It is highly unlikely that if the psychometric tool has high psychometric properties, the report will be inconsistent. In that case, there is a need to break the assumptions and re-align our beliefs so that we can use the report to reduce the blind spots we may have.