Table of Contents:
The construct of emotional intelligence, defined as the ability to perceive, to use, to understand and to regulate emotional information (Mayer & Salovey, 1997; Mayer, Caruso, & Salovey, 2000), has now been lively and controversially discussed for more than 25 years. Despite extensive research, different questions about conceptualization, measurement and validation can only be inconclusively or insufficiently answered. Research questions of this doctoral thesis are therefore: 1. What happens when a test of emotional intelligence is evaluated with different scorings? Which scoring is to be recommended? 2. How does a different response format (“Ticking“ vs. “Allocating“) influence results of an emotional-intelligence-test, the ViTEIP-V2 (Videobasierter Test zur Erfassung der Emotionalen Intelligenz im Pflegeberuf)? 3. Is a video-based presentation appropriate for the ViTEIP-V2?
In the first study, data of N=113 mostly student participants were collected. In the second study, two samples of N=138 and N=122 were recruited from seven nurse's training schools. All participants completed the two performance tests of emotional intelligence ViTEIP-V2 and TEMINT (Schmidt-Atzert & Bühner, 2002) and further tests.
After data evaluation, the author recommends utilization of so-called target-based scorings. Target-based scorings involve a comparison of the participants’ answers and the answer of the target-person who provided information on the situation. Beyond fundamental critical objections referring consensus-scorings (using majority view to generate the correct response-option), higher variance and reliability as well as slightly better validity favor the use of target-based scorings. Three newly developed target-based scorings were able to reduce influence of an existing participant’s response bias.
Use of response profiles as scoring method significantly reduced bias. Due to low reliability of this scoring, the profile evaluation is only conditionally recommended. The parameter “Discriminability index“, deriving from signal detection theory, was used within the scope of a second scoring. It enabled a predominantly bias-free data collection. Splitting items into two parameters “Emotional Sensitivity“ (dealing with existing emotions) and “Emotional Specificity“ (dealing with absent emotions) led to the preferred scoring technique for ViTEIP-V2 and TEMINT. Effects on participants’ answering strategies were minimized and interpretation of correlations concerning validity was much better with two parameters compared to one single parameter. Independent of the scorings, some questions regarding validity of ViTEIP-V2 and TEMINT remain unanswered.
Change of the response format from “Ticking“ to “Allocating“ had an influence on subjects’ responses. A high demanding character of “Allocating“ compared to “Ticking“ led to an increased use of the answering option “no emotion”. As actual performance capabilities were distorted, the usage of “Allocating“ is not recommended. In this study, expected benefits of a video-based version compared to a text-based version of ViTEIP were not supported. If there are no advantages of a video-based test, the author suggests careful consideration of the efforts that the development and administration of such tests require.
By analyzing methodical questions and developing new scorings, the present work contributes to improve measurement of emotional intelligence. Particularly regarding results on the differentiation between “Emotional Sensitivity” and “Emotional Specificity” challenges the assumption of an unidimensional and homogenous construct of emotional intelligence. Further research is suggested, as existing data does not allow systematic analysis of scoring, response format and presentation mode within one sample. To generate future studies in a way where a possible complex relationship of all three factors can be simultaneously analyzed is recommended.