In part 1 (December 2021) we presented the high-level design criteria for a cassudy case study. The following part 2 (February 2022) focuses on more detailed layout aspects of how to present the options to the test subject and how to choose and order criteria to finally identify the test person's traits.
This part (part 3) will add more insights on how to choose the values of attributes/criteria within a cassudy case study, how to evaluate traits for single individuals and for groups of individuals and finally you will get a brief introduction into the concept of „manipulations“ in case study designs.*
*Please make sure you are familiar with the concepts presented in part 1 and 2 of this series in order to be able to get the full benefit from part 3. Please see related articles below.
Setting attribute values to enable a perfect evaluation of the case study data sets
The key data captured in a cassudy study is the attention distribution across all attributes/criteria of an option. The level of attention is defined as a combination of the frequency of accessing the data for a given attribute and the total access time for the same attribute. The access time for a given attribute will depend on the amount of information presented for this attribute (this may be just a simple figure or it may be a statement describing the „value“ of the attribute) and on the time required to comprehend the content for the test subject. In a perfect set of attribute values all attributes should be similar in complexity and comprehension effort to avoid any systematic bias of the attention level.
It is therefore recommendable to use the same type of attribute value presentation and complexity throughout the entire set of attributes for an option within a cassudy case study. So if you decided to use the more complex attribute value approach you may have to artificially complexify attributes that could be expressed in a simple data format.
There are other, more sophisticated approaches possible as well but describing them here would reach beyond the scope of this article. If required, these concepts can be discussed in a feasibility assessment concretely.
Applying cassudy for individuals and/or for groups
So far, we discussed case studies that will be processed by an individual, thereby collecting a data set which is then in turn evaluated and leads to determining the characteristics or traits under investigation for the test subject. Typical business scenarios would be in recruiting, people development or leadership and compliance management.
But especially in larger organizations there is a demand to compare groups of individuals to each other. Let´s take the example of risk propensity again: Is the management team in Asia more risk averse than the one in the Americas? In order to find this out you would run the respective case study with all the individuals of the Asian management team and the Americas management team and you would consolidate the results from the different teams into one combined meta-result per team.
So far this is not a surprise. But what was done implicitly by this course of action is to apply a „manipulation“ to the test. The manipulation is that the test subjects are homogeneous within a group but differ across the groups by their cultural background or the social environment they are embedded in systematically. For the example chosen here we would call the manipulation a „natural“ one as the allocation of the test subject is made by the „natural“ organizational or geographical background and the groups memberships are not defined by other methods.
Manipulations are manifold
For an HR organization in a larger enterprise many manipulations may be of interest when applying cassudy case studies to groups of individuals. The obvious ones are gender, age, hierarchy level, region and functions (core, support, management). But there are certainly more sophisticated manipulations that might give deeper insights into how the enterprise organization works. From a methodological perspective there is no limitation in applying cassudy case studies to any of such manipulation scenarios.
Manipulation „within“ a test subject
A further idea of manipulations shall also briefly be discussed and explained: the so called „within-subject“ manipulation. As we have seen above, a manipulation is a systematic difference in the setup of processing a case study. For groups (see above) we used the example of regional background as the manipulation while all the subjects across both groups still process the exact same case study.
In an „within-subject“ manipulation, the test subject processes a case study with two (or more) decisions to take and the two decisions differ by a single element. Here is an example: the task of the case study is to assess two applicants for a role in an option-wise procedure and the manipulation is that for Applicant A you provide the information that this applicant is an in-house candidate and Applicant B is an external candidate. The manipulation is „internal vs external“ and you would like to check if the assessment process systematically differs for these two different cases.
Again, the manipulation scenarios you may think of are manifold but they all can be processed with the cassudy case study methodology.
If you want to learn more about how cassudy may help to overcome your pain point, feel free to contact me directly:
0176 1022 2375
 A typical approach for a „artificial“ complexity increase is to give ranges instead of single data values or to provide a value in combination with a probability figure
 This wording is taken form academia and describes that the two test sets contain a defined difference (the „manipulation“). The most prominent and well known manipulation happen in clinical test for new pharmaceuticals, where one groups gets the drug with an active substance and the other groups gets placebos