7.2 Key Contributions
7.2.2 The Study
The study run as part of this thesis (and discussed in chapter 6), explores the differences between co-located and virtual teams creativity.
The main finding from this study is that the performance of the treatment group (virtual teams) does not significantly differ from that of the control group (co-located teams). Such a finding enables me to address research question 1: “Do virtual or co-located teams perform creative tasks better?”. This finding is confirmed by a statistical comparison of the treatment and control groups across a range of metrics. This is a significant contribution to both Psychology and Creativity Science because it represents the first empirical comparison of co-located and virtual team creativity.
The Originality scores provide an indication of the novelty of responses. This is the most important metric because creativity is synonymous with the develop- ment of new things. It is therefore highly significant that there is no difference between the Originality scores of co-located and virtual teams across Picture Construction, Picture Completion, Parallel Lines and Alternative Uses activities. Such a finding contradicts many of the opinions expressed by the practitioner and academic communities [175, 176, 177, 178] who argue that virtual working can never perform as well as traditional co-located teams. More importantly this finding serves to support the work started by Girotra et al. [174] that argues that virtual teams are capable of complex socio-cognitive processes equal to co-located teams.
The Elaboration scores provide an indication of the level of detail in the draw- ings, and for the most part these scores demonstrate no discernible difference be- tween co-located and virtual team performance. However, there are two activities during which co-located and virtual team performance differs: the Parallel Lines and Alternative Uses activities. The latter is highlighted by Torrance [8] as po- tentially problematic due to the application of the elaboration scoring technique to text responses. Torrance [8] notes that because of this reason it is often better to treat this as an optional and sometimes unreliable metric. It seems sensible therefore to dismiss the Alternative Uses data as unreliable.
However, the difference in co-located and virtual team Elaboration scores during the Parallel Lines activity remains anomalous. Further analysis has been carried out to try and understand why this difference occurs. First of all the meta data generated by the CreativeTeams tool has been assessed. This data confirms that virtual teams actually use fewer touch interactions and spend less time drawing when completing the Parallel Lines activity than their co-located counterparts. The amount of dialogue used by teams during this activity was then assessed to see if there was any difference in their verbal interactions. The number of utterances used by co-located and virtual teams was compared and no difference found. This suggests that co-located and virtual teams talk just as much despite putting in differing levels of detail into their responses. Furthermore, no difference is found in the number of utterances used when comparing dialogue during the Parallel lines activity with dialogue during the Picture Completion activity. This indicates that teams use the same level of interactions regardless of activity.
These contrary findings provide no clear indication as to why the teams should perform so differently in terms of the level of detail in drawings during the Par- allel Lines activity when the same teams perform so consistently in the other activities. My hypothesis is that this particular activity does not adapt as well to the team environment as the others. This theory is based on my reading of the teams transcripts, during which it becomes apparent that many teams find the activity boring, repeatedly having to generate responses to the same starting shape. Indeed, some teams actually comment “Creativity requires inspiration” during this activity. The majority of teams completing the parallel lines activity also obsess about the need to complete as many of the possible 30 shapes as possible, in contrast with aiming to complete fewer responses but in more detail. Torrance [155] specifically designed this activity to create this tension but it is possible that it is compounded by the team environment. This is one explanation. However it still doesn’t explain why the virtual teams put less detail into their drawings and yet still get the same Originality scores, which are arguably a more important indicator of creative performance.
In conclusion I am therefore unable to attribute a cause to the difference in these elaboration scores. It is an interesting outlier in the broad set of data
that otherwise indicates no difference between co-located and virtual team perfor- mance. This anomaly may be an indication that there are significant differences in performance that aren’t reflected in the other metrics, it may be an indication of a problem with the implementation of this aspect of the tool (although no teams reported it as such), or it may be that the data itself is problematic. This represents a key opportunity for future research to explore whether this problem is repeatable, or to adapt the current activity and see if the trend remains. I would suggest changing the instructions given to teams and the removal of the visible canvas count during the activity as this seems to cause the most tension in teams. It is possible that such a simple tweak may remove the pressure to complete of as many of these drawings as possible that teams exhibit, hopefully increasing the reliability of this measure.
The study therefore results in two important findings. Firstly the study pro- vides a strong indication that co-located and virtual teams perform similarly throughout the majority of the activities in the test. Secondly, there is an un- explained anomaly in the analysis, that co-located teams and virtual teams put significantly different levels of detail into their drawings during the Parallel Lines activity only.
These contributions and findings are significant for a number of fields. First of all, this is the first team study completed using a digitised team form of the Torrance Tests of Creative Thinking [8], and it therefore represents an important contribution to the area of Psychology research concerned with understanding and measuring creativity. In particular it builds on the work of Kim [9] who critiques the TTCT, ultimately concluding that it remains relevant but will require adap- tation. Secondly, this work provides an important trigger for dialogue within the management and practitioner community about what virtual teams can achieve. In particular it provides a repeatable quantitative analysis of team performance in the two main work environments typically employed. Finally, this contribution should have a meaningful impact in the computer science community, providing an indication that virtual teams are capable of complex socio-cognitive processes above and beyond arduous development work.