Operationalization: Making a Big Concept Measurable

A big word, but really, all it means is how we express the phenomena we are looking to explore. For instance, if you wanted to study religiosity, you might operationalize this as “attending church services once a week.” There can be many observables to explain a phenomenon under study, this is just one possibility.

With its Self-Esteem Weekend, Dove will be interested in evaluating its success. They may do this, as I discussed in my previous blog  by survey response. While I discussed sampling first, creating the research design would, in practice, have come first.

So Dove might decide that they are interested to study the level of girls self-esteem. There are many ways to capture this data. Today, I will chose to capture the level of self-esteem using an index. There are a few kinds of indexes. In an index, each variable being measured is given equal weight. So, I might ask questions like:

1.       How happy do you feel at school?

Happy  1(very)    2 somewhat)    3(neither)    4(somewhat)    5(very)  Unhappy

2.       Which describes you?

Pretty  1(very)    2 somewhat)    3(neither)    4(somewhat)    5(very)  Ugly

3.       Do you feel comfortable expressing yourself to your friends?

Comfortable  1(very)    2 somewhat)    3(neither)    4(somewhat)    5(very)  Uncomfortable

And so on…

In these few examples, I have operationalized self-esteem as measurements of whether a girl feels happy at school, thinks that she is pretty, and feels comfortable expressing herself to friends. These are a few of many ways to operationalize a very complex concept. On an index, these questions measure responses and weigh the value of responses equally.

So if one girl chose her answers, respectively, as 4, 3, 5 we might say she scored 12 points on the self-esteem index. Another girl whose responses were 2, 2, 1 would score 5 on the self-esteem index. This very clearly showed that first girl scored higher on the self-esteem index. If our operational measurements are valid and reliable, then we can conclude that girl 1 has more self-esteem than girl 2.

(To measure whether girl’s enjoyed the Self-Esteem Weekend, we might do something similar as above, but use questions that measure whether the girl’s felt they “learned a lot,” “had fun,” or “would like to do it again next year.”)

So, today we operationalized a concept (we came up with a way to measure or observe a phenomenon) and learned how to put them on an index.

Along with our topic of Dove research, appropriately, Dove has posted to their Facebook wall today that they conducted their 2010 research on women and self-esteem, and concluded that “women still don’t feel comfortable describing themselves as beautiful.”  Check out the results.