This blog post is a reflection on current trends in personality assessment. Broadly speaking, it’s aimed at non-psychologist colleagues and HR contacts who have an interest in personality measurement but who may be confused by what’s on offer. Share this post by all means, but please only link to the original. Please do not copy or re-post without full credit.
For those of you who have even a passing interest in workplace personality assessment, you may have come across a new word recently: Nipsative. It’s a sort of ‘combination approach‘ to personality assessment (combining the two words Normative and Ipsative). In all honesty, I rolled my eyes when I first heard the word, but as someone who assesses character and temperament as a matter of routine, I was duty bound to explore and evaluate…
So what exactly is ‘Nipsative’?
To answer that question, you have to start with two distinct branches of modern personality assessment. I’m talking about questionnaires such as:
- Cattell’s 16PF, the 15 Factor Questionnaire, the California Psychological Inventory… These questionnaires are Normative. You need to understand Normative assessment first.
- DiSC / Personal Profile Analysis, Insights… These fall under the umbrella of Ipsative assessment.
Then… Finally… If you’re ready for it… You can tackle the new kid on the block: Nipsative.
Questionnaires such as 16PF and 15FQ+ are Normative by design, meaning that results from different people can be directly compared. Individuals who complete Normative questionnaires are measured in precisely the same way (ie: they answer exactly the same set of questions), and the results can be presented on standardised scales. You can think of it a bit like using a tape measure.
For example, if one person has a standardised score of ‘2/10’ on an Introversion-Extraversion scale and another person has a score of ‘8/10’ on the same scale, you can be confident that one is more Introverted / Extraverted than the other. We are using the same dimension, measuring it in the same way and presenting results on the same scale, just as though we are measuring height in metres and centimetres. In personality assessment, this means that…
The traits being measured are independent of each other, so a respondent can score high on all scales, low on all scales, or any high-low combination. This is an important point to note going forward.
Before we go any further, let me give you some Normative-type questions:
- ‘Do you have Apples?’
- ‘Do you have Red Apples?’
- ‘Do you have Green Apples?’
- ‘Do you have black-and-white striped Apples?’
These are straightforward question and I’d like you to give me straightforward answers. Honest answers, please. You might answer ‘Yes‘, or ‘No‘ or ‘Not sure‘ to any or all of the questions. I don’t mind how you answer; just be honest.
So now let’s imagine that I also have answers (to exactly the same set of questions) from another 1000 people. Now I can begin to work out what your answers actually mean. I can calculate an average, for example, and I can work out whether you have more (or fewer) Apples than the average person. It starts to give real meaning to your answer; I can begin to infer something.
There are enormous benefits to Normative assessment, and I’ll return to them later, but for the moment I’ll deal with some possible drawbacks.
Firstly, despite the benefits of Normative testing, there has been one persistent criticism, and that revolves around the issue of ‘faking’. That is, some people argue that some people are better at spotting desirable answers in questionnaires, thus introducing potential distortion into the measurement process.
For example, you might have told me that you had a whole load of Apples, including some black-and-white striped ones, when in reality you only had three small green ones. When responding to my questions, you ‘faked good’ (or just plain lied), for whatever reason. Don’t worry; I’m not judging you. Maybe you thought I would look on you more favourably if you had a ton of multi-coloured Apples?
Of course, as you probably know, when it comes to actual personality assessment, high-quality test publishers take active steps to identify and control for the possibility of faking good (PS: I know for sure that you don’t have a black-and-white striped apple!). In any case, it is perfectly possible (indeed, advisable) to validate results following the assessment process (through structured / behavioural interview, via reference check, etc.). I might, for example, ask you to prove that you really had so many Apples. I might want to see evidence of them, especially the black-and-white striped ones, or perhaps speak to someone who could support your claim.
But that’s not quite all.
In the ‘Apples’ example, we’re talking about something tangible. An apple is an apple after all. In real-world personality assessment, we’re measuring something less obvious. Who’s to say, for example, whether watching the news is more or less desirable…? Is it better to use telephone or email? Although you might think some responses are ‘obvious’, it’s not so easy in practise.
There’s also another issue to consider. Research has suggested that ‘faking good’ in personality assessment may not actually be a bad thing, nor even a conscious act. Indeed, there is a suggestion that those who can spot socially desirable traits in a questionnaire are more likely to recognise socially desirable behaviour in the real world. They may not be deliberately ‘faking’; they may simply be far more effective at recognising what is required in a given scenario. And if they can do that in a questionnaire, they are surely more likely to do it in practice. They know how to behave.
In any event, a great deal of work goes into Normative questionnaires: To ensure the questions are fair; to validate the scales; to prove that questions measure what they are supposed to; to check for consistency, reliability, faking good (and sometimes faking bad); and ultimately to build up a picture of an individual relative to other people. Normative questionnaires are straightforward measures designed for comparing people. It’s just a more objective way of doing what we do as human beings.
I want to touch briefly on one other issue relating to Normative questionnaires, and one that is very rarely spoken about.
There are many practitioners – psychologists and non-psychologists alike – who are uncomfortable with the idea of comparing people. They seem to dislike giving feedback which might be ‘harsh’. They don’t like explaining to a candidate that a particular trait makes them more or less effective at something; more or less ‘suitable’. For these people, the measurement process of Normative testing is too sharp and objective. They are far happier when they can provide feedback which is gentle and non-threatening, where everyone is a ‘winner’.
And so now we can introduce the…
Ipsative tests have a long history and appear to have an advantage over Normative tests when it comes to ‘faking’. They also allow practitioners to adopt a softer approach during the feedback process, focusing on relative strengths within the individual (ie: rather than differences between people).
Very often, Ipsative questionnaires focus on TYPES rather than TRAITS, and certainly for things like Development and Team-Building, they can be extremely valuable. They don’t threaten anyone, because EVERY ‘type’ has strengths; no ‘type’ is any better than another.
The design of questions in Ipsative assessment, however, is fundamentally different to that of Normative questionnaires, in that respondents are forced to choose between different options, all of which can appear favourable.
In effect, the questions posed are not:
- ‘Do you have any Apples?’ (Normative)
But more along the lines of:
- ‘Which do you have more of: Apples or Oranges?’ (Ipsative)
Okay, I accept that I’m simplifying things a little, but you get the idea. Basically, the ‘forced choice’ element in Ipsative assessment makes it all but impossible to manipulate or fake answers (and most would argue there is no benefit to even trying).
But there is a problem with this type of question; a VERY BIG problem…
When answering ipsative-type questions, you can’t have both Apples and Oranges. One fruit is always at the expense of the other. And this results in a fundamental drawback: The things being measured are interdependent, and it is no longer possible to draw comparisons.
With Ipsative personality assessment, the results generated are self-referenced. It’s vital to understand this, so I’ll say it again. The results are self-referenced. They aren’t relative to anyone else at all.
Regardless of the lack of distortion, results from Ipsative assessment cannot tell you whether one candidate has more Apples than another candidate. You can’t say with certainty that one person who appears to have Oranges has any more or less of that fruit than the average person. All you can say is that this particular person has more (or fewer) Apples than they have Oranges.
What’s more, every time a respondent agrees to having Apples, he/she is potentially excluding the possibility of having Oranges. They may still have more Oranges than the average person, but the format of the questionnaire prevents you from making the comparison.
If you’re asking yourself why this matters, then I would argue that you haven’t fully considered the uses of personality instruments in the workplace. They are important and valuable tools in identifying strengths, qualities and behaviour, and we tend to use them in two distinct ways: SELECTION and DEVELOPMENT
In Developmental scenarios, we are usually looking to identify areas of relative weakness, and then work with the individual to either change behaviour or encourage them to compensate in some way for limitations. Ipsative questionnaires can be very useful in these circumstances, especially if the person involved is in ‘defensive’ mode. We can explain during feedback that every individual, almost by necessity, will have high scores in some areas and low scores in others. It doesn’t matter that we can’t compare them with anyone else, because we don’t need to. We only need to pick out areas for development.
In Selection scenarios, by contrast, we are almost always looking for the ‘best fit’ for the role. I would even argue that it is necessary to compare people in order to do this. We do it instinctively anyway. It’s one of the reasons why most organisations these days have ‘structured interviews’… to help with the comparison. What structured interviews aim to do is standardise the process. We still accept that people fake good in interviews, but it doesn’t stop us from standardising the process where possible. It makes perfect sense to do exactly the same with personality assessment.
We also draw comparisons when evaluating CVs, albeit in a less analytical way. We compare relative experience, educational background, qualifications, track record, etc.. Let’s be honest, we’re designed and built to compare.
To me, it’s as plain as day… If we want to select the best Apple, we have to compare our choice with other Apples. For me, Normative assessment hands-down beats Ipsative assessment in selection scenarios. I could also argue for the value of Normative assessment in Developmental contexts, but I’ll save that for another day. After all, we’re here to shed some light on the new kid on the block: Nipsative.
So here we go…
Okay, now that we’ve covered Normative and Ipsative assessments, we can move on to explore Nipsative assessment, the ‘newest format’ personality questionnaire. Remember this is a new word, combining Normative and Ipsative.
The Nipsative format seems to be a brave attempt to combat the weaknesses of purely Ipsative assessment by adding another step in the questionnaire design. I’ll use the same kind of examples to try and explain…
First of all, remind yourself of these example questions from earlier…
- ‘Do you have Apples?’ (Normative)
- ‘Which do you have more of: Apples or Oranges?’ (Ipsative)
In Nipsative questionnaires we have this:
- ‘Do you have Apples?’ (Normative)
- ‘Do you have Oranges?’ (Normative)
- ‘Do you have Lemons?’ (Normative)
- …and then…
- If the answer to any of these questions is the same (whether the answer is ‘Yes, ‘No’ or ‘Not Sure’),
- ‘Which do you have more of…
- …Apples or Oranges, (Ipsative)
- ……Oranges or Lemons, (Ipsative)
- ………Lemons or Apples? (Ipsative)
Does this seem like a step forward in personality assessment? If it is, I’m struggling to see how.
Firstly, we’ve re-introduced the possibility of distortion (the thing that Ipsative questionnaires purport to avoid). Secondly, we’ve still got the lack of independence between scales. And third, there seems to be a risk that candidates will adopt differing response styles. In particular, there will be a whole tranche of people who either:
- Treat the entire questionnaire as Ipsative, or
- After treating one or two questions as Normative format, realise what the game is and resign themselves to Ipsative, eliminating the post-normative, forced-choice stage altogether*.
*For example, rather than saying they have lots of Apples, lots of Oranges and lots of Lemons, and then realising that they are forced to separate their answers, they immediately answer: Lots of Apples, quite a few Oranges and not many Lemons. They jump straight to the Ipsative format to save time and bother. This may be especially the case with people who are short of time or who lack patience!
I don’t want you to leave this article thinking that Ipsative / Nipsative questionnaires have no value. Quite the contrary. In fact, I know that in developmental scenarios they have the potential to start some really deep and informed conversations. I’ve used such results myself to great effect.
But are they the best tool to use in selection assignments?
I stand by my earlier assertion that when we are selecting people to meet the demands of a specific work role, we are trying to find the best person for the job. And if we are looking for ‘the best’, we necessarily need to make direct comparisons with others who may not be the best.
Have you ever picked an Apple from the fruit bowl and inspected it? Of course you have.
And have you ever put it back after inspection and chosen another? Again, of course you have. Because you’re looking for the best Apple. And even if there’s only ONE Apple in the bowl, you will compare it with your idea of a good Apple. As humans, we compare all the time, and when it comes to personality assessment, you can only do that fairly with Normative assessment.
Sadly, I’ve seen people compare Ipsative profiles. I’ve seen them try to explain why one person’s scores are better than another: ‘We need Oranges for this job and this person has the most Oranges.‘ It comes from misunderstanding the measurement process, and I’m sure by now you can see why… Because it’s in our nature to try and compare. We do it everyday. Whenever you describe a person…
.. ‘He’s really quiet.’
…. ‘She’s very friendly.’
…….. ‘He’s a worrier.’
……………. ‘She’s so astute.’
… you are comparing the person against your idea of what is normal; what is average; what is typical. What you’re actually saying is: ‘He’s much quieter than the average person’, ‘She’s far friendlier than most..’, etc..
It’s a language we all understand. We make sense of the world – and of people – through comparison.
In Praise of Normative
I’m not dismissing the criticisms of Normative testing. Faking good happens. But it usually happens in places that practitioners know about: Warmth, Friendliness, Trust, etc… the desirable qualities. And knowing that, we can interpret results accordingly; we’re not so easily taken in. Anyone with experience of Normative results knows what to look for; the signs are usually recognisable.
I would also like to suggest that some test developers are overly keen to highlight the problem of distortion in normative assessment; sometimes the ‘red flags’ are raised unnecessarily. Consider my own experience…
I’ve been completing personality questionnaires for 25 years. I’ve tried to complete them honestly and accurately – I really have tried – but I have never, EVER, EVER managed to deliver a set of results which does not show some level of distortion. No matter how hard I try, results show me as faking good. And yet I can recognise myself in results, and people who know me well can see me too (yes, warts and all).
So, I’ll leave you with this…
The overriding strength of Normative assessment is its ability to compare people fairly. It may not be a perfect measurement process but – particularly when it comes to selection assignments – it’s far and away the best process for complementing what we already do.
So go ahead and pick your Apple. I bet you want the best.
“There’s small choice in rotten apples.” – William Shakespeare, Taming of the Shrew