So operationalizing the measures was
an interesting process and it was fun for me to see how the sausage was made.
You know, I knew a fair amount about psychological research,
but it wasn't my primary work.
And seeing how questions would evolve,
but also how to get the measures to be more authentic,
how to get them to really measure these human differences.
So interviews and surveys are fine,
but they are fraught with the potential for bias,
reference bias and social desirability.
People looking at these questionnaires know, oh,
that's the better answer;
that reflects better on me.
And so our intuitions,
which were, I guess initially,
that if you asked someone about themselves and they were
assured that the research was anonymous that they
would truthfully respond and that that would give you
a real insight into how the real world worked.
But it's not that simple and so it's not that easy to measure.
Really, one needs a little bit of stealth in getting this to work.
So getting the hypothesis and moving it into different kinds of behavior measures, these,
what evolved to be called the academic diligence task,
which would measure the ability to sort of stay in a boring,
tedious math task rather than going on to
the other side of the screen where there's this temptations,
video games and music videos and cute animals.
What we learned was that this was
a very accurate way to measure these so-called non-cognitive skills
because what it allowed for was to see
an ecological picture of how the individual functions.
And that's always a challenge.
You know, the way people actually conduct themselves in school and you know,
we had data about that.
We had demographic data from school administrators and we had a sense
of things like lateness and attendance and grades and things like that.
But now we were able to see how those things correlated with
real-time behavior and to perhaps get
a better insight in ways to manipulate those qualities.
So, over the course of the research,
we came to discover that
individuals could be altered.
There could be an intervention.
We could help them be their best selves by helping them to know
themselves better and getting specific strategies for growth and change.
So the methods and measures we've
used have evolved over the course of the five years that I've been at the lab.
They started with very simple paper and pencil surveys and then these days we do
online distance measures that can really
look at a variety of individuals in different settings and to get a better sense.
And it's a much more eclectic collection of measures than when we started,
so that there are now behavioral and other kinds of ways of looking at human behavior.
And they're just so much more valid,
now that we have a better collection of instruments to measure these behaviors.
Claire Robertson-Kraft, Ph.D.Director, ImpactED
Angela Duckworth, Ph.D.Christopher H. Browne Distinguished Professor of Psychology
