By C.H.I. Talent Assessment, Mar 13 2017 10:20AM
In the summer of 2016 new Sat tests were implemented in schools across England. The Sats are tests taken by 11 year olds to track their academic progress. When the first results were returned, some local authorities had their percentages reaching the expected standard drop by almost 40%.
So, what does this have to do with the candidate assessment process? What it does, is highlights the problems that we have with expectations in assessment. To demonstrate this, I will use the example of the Sefton local authority. This region saw a drop from 81% reaching the previously expected standard to 56%. There are two possibilities of why this might have occurred-
A) Last year, 30% more children failed to reached the expected standard.
B) The expectations set for the test were incorrect - the goal posts were moved.
In 2014, Sefton saw an 80% reaching expected standards, in 2013 78%, and 2012 76%. Whilst not massive, there was an increase year on year for the past 4 years, suggesting that the area was gradually improving (a similar pattern could be seen on a national scale, with 80% reaching the expected level 4 and above in 2015, 78% in 2014, and 75% in 2013 and 2012). For that figure to then immediately drop to 56% would require a massive reversal in the trend. More likely is that the change in test was not followed by a change in expectations.
If Sefton’s results were taken in isolation, then they would be highly worrying. If there was no external reason for the dip in the number of students reaching expectations, then you could expect overhauls of the schools in the region. A change in approach to how the children were being taught, along with an investigation into the quality of teachers who are teaching in the region, would be expected. Fortunately, we have access to the national data as a comparison group, and can see that the standards set for the new tests were not reached by almost every region in the country, so Sefton need not panic just yet.
Consider your business, and how you’re currently assessing your job applicants. How can you check your standards? Unlike the Sats, typical job interviews and CV checks don’t have national statistics to compare your candidates’ results to, and many psychometric tools do not have the proper comparison groups. If the tool you're using doesn't have a comparison group that represents typical applicants to the role, you can't trust that the findings of the assessment are accurate. When you can’t trust the results of an assessment, there’s no point in doing it at all.