Cause-Effect Separation and Selection of Case Studies

February 26, 2007

Although there is no consensus in literature on what exactly “success” is for a particular policy, intervention or centre (economic turnover? social integration? Resident/Customer/Stakeholder fulfilment?), one should still be able to put forward, in the freedom of their own point of view, what they see success as, and then investigate the dynamics towards this understanding of success. But I must warn that I will not focus on what success is for us here, our view on the successful suburban centre deserves a separate dedicated thread in the blog. What I want to focus on is how to go about researching into the dynamics of it.

Apart from the big question mark on definition of success, one issue in generally in all discussions about the “success” of the town centre, is that a) signs and b) reasons of success are not separated clearly. One trying to tackle an issue like success will benefit greatly from differentiating which aspects effectively tell them a particular centre is successful, and which aspects they see as contributors/factors helping that success result to come out. Only if we decide on results (effect), we are able to test the possible reasons (causes), to see if they contribute to that result or not.

For instance, let’s say our expected readable result of success is diversity of uses for a moment. This can be made measurable (having existence of group x, group y, group z above a threshold we define, although how to measure diversity is still a discussable issue) and we would end up with centres scoring well or poorly on this indicator. Then we should have a set of candidate set of measurable causes that we think contributes positively towards this diversity result, considering the micro-scale of our project. Again go with the example for a moment, we could ponder on the possible micro-spatial causes that would lead to diversity: size of units, grid size, and variety of micro-spatial patterns in the centre. Then we can test each of these candidate micro-spatial causes for centres that scored well and poorly in terms of “diversity”, were they indeed a factor?

A similar example can be given for “accessibility”. I personally see accessibility as a reason for success, not a sign of success. The sign of success I expect it to contribute towards is the existence of people across the daily time slots, in other words, the continuity of activity across time (again, something we agree on as a team to do with success, but to be discussed under the success title). Now the question of whether accessibility contributes towards this can be answered by deriving our measures of accessibility on micro and/or macro scale on one hand, having a measure for existence/continuity of activity on the other, and see if there is a relationship between the score of these measures in the centres we are working on.

Moving on to Selection of Case studies, trying to put these expected, assumed signs of success forward should not be a step for selecting the case studies. In fact, case studies should be selected independently from what we assume success is. The reason we are selecting a number of case studies to work on in more detail is that we simply can not run our detailed analyses everywhere due to time and funding constraints. However this does not necessarily mean this selection should be random, because we still would like to compare like with like, we want to know what kind of centres, within the macro framework, we are dealing with in the later stages.

The possible case study selection factors we discussed today in our internal meeting, such as: a) size/scale of centre, b) having a residential background, are possible factors to enable us to define our boundary, regarding where- within what type of geography- we are searching for success.

Such simple approach for selecting the case studies would also protect us from the danger of biasing ourselves towards expected success from the beginning; both successful and unsuccessful cases would have equal chance of being included in our selection.

Advertisements

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: