Building causal models

When we are analysing data we might observe that there are associations between elements in the data. The interest is usually to see whether the associations can be integrated into a meaningful explanatory model. To do this it is necessary to apply some heuristics.

The first heuristic involves ordering the model temporally. The questions to be answered then become:

• Which variables or events of those found relevant occur first in time?

• Which occur along the way during implementation?

• Which might be seen as early and late outcomes?


The second heuristic involves considering which variables might reasonably be expected to have a direct impact on other variables, both preceding them in time and having a plausible direct connection. This rule means that we need to think of the mechanisms that could be involved.

The third rule is to check case informants' explanations. What causal linkages do they claim are present? We need to examine our transcripts and fieldnotes very carefully.

The fourth rule is to consider what available research and theories have to say about causal connections.

Using these rules, we have to play with the data for a while. It can help if we put variables or events on cards and move them around into various configurations, looking at the connections that seem sensible. Graphics packages such as Inspiration5 can also be helpful. Another special program that can help develop causal models is COPE (cognitive policy evaluation).6 While primarily designed to produce cognitive maps, it enables us to map in two dimensions explanations and consequences of events through a graphic user interface. Since it also has the facility to trace paths throughout the complex network structures that can be constructed, it can also help to establish cause-effect relationships and hence build causal models.


Marketing researchers undertaking focus group discussions with groups of patients suffering from a specific disease might discover that in severe cases of loss of appetite, patients are treated in different ways by the GP, depending on whether or not the GP feels they can treat the particular symptom. Where the GP feels confident, a proprietary brand of medicine is prescribed as treatment. Where the GP is unsure, referral is made to the consultant at the hospital, who prescribes a non-proprietary brand. The whole process is captured in Figure 11.6.

Poor health


Can treat

Online Survey Champion

Online Survey Champion

There are people all over the world trying to find ways to make money online. From stay at home moms looking to make a few extra dollars to college students and entrepreneurs, the allure of making your own hours and working from home or from the local coffee shop is very appealing.

Get My Free Ebook

Post a comment