5.2 Discussion
5.2.1 Interpretation of results
As the outcome of this research is that the hypothesis is not met and academic subculture does not influence researchers’ intention to start a spin-off, it is inter- esting to think about possible clarifications. First of all, it may be possible that the theoretical model as discussed in chapter2does not fit the situation. As stated in that same chapter, I have made the choice to see culture as a background factor. Nevertheless, there was also the possibility of seeing culture as factor that influences perceived behavioural control instead of seeing it as background factor. Looking at the results, it seems that culture only influences perceived behavioural control marginal significant, but it does not influence the other vari- ables. Therefore, it may be that culture should have placed at the other side of the model, influencing perceived behavioural control only. This could clarify why the outcome is that culture influences only perceived behavioural control. A second clarification for the results is that other background factors are influenc- ing the outcome. In this research I used culture as only background factor and have tried to keep other background factors as similar as possible. Nevertheless, it is very difficult to keep all factors stable; it may be possible that other factors have influenced the intention indirectly. Gender is an example of such a back- ground factor. As visible in chapter4, results, the intention to start a spin-off is higher among respondents of EWI than among respondents of BMS. Neverthe- less, the same chapter has shown that men also have more intentions to start a spin-off than women and that EWI has significant more men than women in com- parison to BMS. It thus is questionable whether the difference in intention to start a spin-off comes from academic subculture, or that is comes from difference in gender. These two factors can also be connected with each other; maybe gender influences culture, which makes both culture and gender different and influence the intention to start a spin-off. Nevertheless, gender can also be an important background factor influencing researchers’ intentions to start a spin-off.
The ignorance of what spin-offs are among respondents can be a third clarifica- tion of the results. One respondent has send an e-mail in response to the ques- tionnaire in which she stated that she was not familiar with the concept spin-off. She had no experience with it and neither had some of her colleagues. This may explain the low response rate among some departments: within one department there were many respondents who did start but did not complete the question- naire. Overall it can thus be that respondents were not familiar enough with the concept ‘spin-off’, as a result of which questionnaires were not always filled in as desired.
Another possible explanation is the inaccuracy of the operationalization. In this case, the theory may be right but the concepts are not operationalized properly. The variables measured then do not correspond with the theoretical concepts and thus do not measure the right factors. For example, a different operationalization of the concept attitude may have lead to a different outcome, more in line with the hypothesis.
As stated in the description of general results (chapter3), 25% of the respondents filled in that his/her function is ‘other’. This is a last reason that may clarify the results as given. As these persons are not academic staff as mentioned by the other categories, it may be that they are supporting staff or students graduating at that department. These persons are less in a position to start a spin-off: sup- porting staff does not do scientific research and students are busy getting their degree. Therefore, they may influence the results in such way that they may have positive attitude, perceived norm en perceived behavioural control, and see the departmental culture as entrepreneurial, but do not have the intention to start a spin-off because they do not have the possibility to start a spin-off.
By now four important clarifications for the results as given have been explained. As there are also limitations resulting from the research design that have influ- enced the outcome, I will hereafter explain these limitations.
5.2.2
Limitations
As this research has approached respondents of only one university, the Univer- sity of Twente, the range of the research is limited. The results are not generalis- able for other universities. Within the university, only four departments have been researched. This is also a limitation for the generalisability of the outcome of the research; it may be that the variables are seen differently at other departments. Besides, there are only two departments per sort of science: two for the techni- cal sciences and two for the social. It is questionable of these departments are representative for the whole technical and social population within the university. As a result of the small range of the research, the sample taken is also very lim- ited. There are only 36 respondents in this quantitative research. As a result of this small sample, results are difficult to analyze statistically as outcomes are barely significant. For example, the culture per function has been analyzed. The outcome is significant (Chi2 = 10,143;df = 4;p = 0,038) but 80% of the cells have an expected count that is less than five of which four expected counts are
below one which means that the conditions for the Chi-square Test are not met. In many other cases the outcome is not significant at all. The small amount of respondents thus prohibits from being able to draw accurate conclusions.