Team production in a field experiment: study of aggregative vs. individual cultural activities

Abstract

This study aims to determine whether team production among peer groups culturally aggregated can affect the individual student performance in an Italian university. Two further hypotheses were tested in order to verify whether students perform better in team than individually and if students culturally aggregated in team perform better than students differently aggregated. By means of a field experiment, 162 students were involved in the study. The participants had the opportunity to sit a new examination model after responding to a self-reported questionnaire. The new model contemplated that the students performed in team for the first part of the program and carried out the remaining part of examination individually. The groups were composed of individuals linked by homogeneous preferences about cultural activities. The results show that team performance affects positively the final score of exam and the students culturally aggregated pass the exam more than students not culturally aggregated. The aggregative cultural activities influenced the test score more than individual effort or other personal characteristics. We therefore recommend the implementation of a new examination models based on cooperation among students supporting the aggregative cultural activities where sharing action prevails as predictors.

Appendices

Appendix 1. Definition, label, and scale of variables

In Table 7, we present the variables of interest with a synthetic definition and relative scale.

Table 7 Variable definition

Appendix 2. Descriptive statistics across courses

In Table 8, we report the descriptive statistics relating to the courses. A similar composition across 12 items of UNESCO questionnaire is shown with respect to Table 1. It is evident that the only changed characteristic is the age. In fact, as we will see later, Age represents a not determinant variable for our study.

Table 8 Descriptive statistics across courses

Appendix 3. Backward linear regression model on Grade Team Part

Starting from the model (8) of Table 5, we run a linear regression model where the Grade Team Part is explained by Culturally Aggregated Team and by only variables in the vectors X_i, Y_i, Z_i, P_i, C_i, and PC_i that do not drop, 1 at a time, to the specification to be significant at least 10%. Shortly, the model resulting is so specified:

$$ \mathrm{Grade}\ {\mathrm{Part}\ \mathrm{Team}}_{\mathrm{i}}=\genfrac{}{}{0pt}{}{3.993\ast \ast \ast }{(1.503)}+\genfrac{}{}{0pt}{}{1.706\ast \ast \ast }{(0.371)}\mathrm{Culturally}\ {\mathrm{Aggregated}\ \mathrm{Team}}_{\mathrm{i}}+\genfrac{}{}{0pt}{}{0.056\ast \ast \ast }{(0.016)}\mathrm{High}\ {\mathrm{School}\ \mathrm{Grade}}_{\mathrm{i}}+\genfrac{}{}{0pt}{}{0.529\ast }{(0.274)}\mathrm{Play}\ {\mathrm{an}\ \mathrm{instrument}}_{\mathrm{i}}+\genfrac{}{}{0pt}{}{0.591\ast \ast }{(0.254)}{\mathrm{Political}}_{\mathrm{i}}-\genfrac{}{}{0pt}{}{0.514\ast \ast }{(0.242)}{\mathrm{Conference}}_{\mathrm{i}}+\genfrac{}{}{0pt}{}{0.350\ast }{(0.207)}{\mathrm{News}}_{\mathrm{i}}+{\varepsilon}_{\mathrm{i}} $$

with adjusted R-squared = 0.223, observations = 162, F = 8.684 p(F) < 0.01. Estimation results are sensibly greater than the ones of the full model and now, News participates at level 10% with a positive magnitude of + 0.350 points to explain Grade Team Part.

Appendix 4. Evaluating cultural connection across age and passing the intermediate team test

To exclude potential effects of participants not commonly aged of university study, we focus on the reduced sample of 155 students aged under 28 years old.

Table 9 Cultural aggregations and team performance. OLS and Tobit estimates for only students aged less under 30 years old