A Machine Learning, Artificial Intelligence Approach to Institutional Effectiveness in Higher Education by Moye John N.;

Author:Moye, John N.;
Language: eng
Format: epub
Publisher: Emerald Publishing Limited
Published: 2019-05-16T00:00:00+00:00

Aggregating Functional Data into Causal Triads

As has been discussed, single indicators are inadequate to demonstrate the complexities of institutional performance. In this example, there are 90 performance indicators, suggesting that any single indicator only explains 1/90th of the total performance information. The indicators are triangulated into meaningful groups to provide practical meaning. In this assessment system, the KPIs are clustered into triads to demonstrate performance. The correlations between the KPIs define these triads based on the strength of the correlations. These triads demonstrate the interrelationships between the indicators, which represent the synergies of functional performance (Nagarajan & Maluk Mohamed, 2017).

A subset of the data is analyzed to calculate the KPIs between indicators and demonstrate the procedure with all 90 indicators. Table 65 presents a correlation matrix reported from a factor analysis conducted with the subset of 14 indicators selected from the 90 indicators contained in the original functional data set.

Table 65 presents the correlations between each of the 14 indicators and each of the other 14s indicator. With this information, it is possible to identify the strongest indicators that are varying together or “covarying.” A correlation below –0.400 and above +0.400 is considered a strong correlation, while those between –0.400 and +0.400 are considered weak correlations. A correlation of or around zero indicates no correlation. The strongest correlations become the KPIs of institutional performance because they represent the most powerful drivers of institutional performance.

Each KPI with its two strongest correlates (positive or negative) triangulates into clusters of three KPIs or KPI triads. Each triad contains the primary indicator and its two highest correlating indicators either positive or negative. The clustering procedure is a method of calculating and revealing the synergies of functional performance (Drouin, Stewart, & Van Gorder, 2015; Tewari, Singh, & Tewari, 2016).

Table 66 presents the results of the clustering of the KPIs calculated in the correlation matrix reported above. Those correlations presented in bold are positive correlations, and those in italics are negative correlations. The correlation for each indicator and a description of the relationship is included to translate the meaning of the cluster into functional performance.

Table 66 identifies the two strongest correlations for each indicator. Some of the resulting triads contain strong correlations; others do not. When used to describe institutional performance, the data model excludes triads with weak correlations from the data model. In the systematic assessment of performance, this analysis creates information that program staff and managers use to determine the causes of actual functional performance.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.