Comparison of Statistical Methods for the Analysis of Affective Haptic Expressions

Several studies were conducted to show the relevance of haptics for conveying emotions to users. These studies usually cover recognition rate of emotions from haptic expressions. Surprisingly, the analysis of features of these haptic expressions has been in counterpart often limited to a classical analysis of variance. This method is limited since it can neither highlight multiple possible expressions of a given emotion nor compare several emotions or features simultaneously. This paper presents a methodological approach for collecting and analyzing haptic expressions of emotions. We compare three statistical methods, namely analysis of variance, principal component analysis, and clustering. Over this study we will highlight the advantages and drawbacks of each method for the analysis of haptic expressions of emotions.