3.2. Variables, Structure of Data, and Description of Statistical Methods
This study examined the effectiveness of instruments supporting inter-organizational cooperation in RES markets. The effectiveness was measured with two variables: the number of meetings during events and the number of established partnerships initiated as a result of these meetings. These two factors were our dependent variables. Our independent variables were three characteristics of the events, namely, the location of the event, type of event (brokerage event versus company mission), and type of leading partner (public versus private).
The first stage of statistical analysis was divided into two parts. In the first part, the numeric data regarding the meetings, which took place as part of the promotional events recorded in the EEN database, were analyzed. On the other hand, the second part of the study included data on the number of partnerships established between the entities participating in these events.
The preliminary results of the analysis of the number of meetings taking place within the studied cases lead to the conclusion that no meeting took place during a significant part of the events. Out of 304 analyzed events, no meetings were recorded in over 86 cases. Thus, at every fourth promotional event, on average, there was no meeting. This may be due to the fact that the organizer of the leading event developed the program independently and did not propose cooperation to the EEN.
To approximate the structure of the number of meetings held as part of individual promotional events registered in the EEN database,
Table 1 presents basic descriptive statistics for events from all over Europe in total.
When analyzing the values of the above statistics, a very large difference can be noticed between the average value of the number of meetings and the median. Statistically, there were approx. 137 meetings per one of the 304 analyzed events, while the median for the number of meetings was much lower, approx. 55.5. This conclusion is confirmed by the value of the coefficient of variation, which is over 176%, indicating a significant variation in the number of meetings. This fact is additionally confirmed by the value of the range. The difference between the highest and lowest number of matches is 1446. The values of the above statistics may point to high outliers. The results obtained were also significantly impacted by the occurrence of a large group of events during which no meetings were recorded. In conclusion, in the part of the analysis concerning the number of meetings, attention is drawn to the fact that there is a large variation in the number of meetings for individual events. The results of the above analysis and the significant variation in the tested values may lead to the formulation of hypotheses regarding the factors that differentiate the number of meetings during the studied events.
When analyzing the data on the declared cooperation following a promotional event, it can also be noticed that there were a significant number of events that did not develop into any cooperation. Out of 304 analyzed events, in 38% of cases, no cooperation was recorded. Thus, the participants of three out of five promotional events, on average, showed a willingness to cooperate.
Determining the basic descriptive statistics for the number of partnerships recorded in the EEN database would help to better characterize this variable. The obtained results (see
Table 2), similarly to the analysis of the meetings, indicate a large diversity of the studied set.
The first symptom is a significant difference between the mean and the median. The analysis shows that, on average, a promotional event generated approx. 32 business connections. However, the median for this value is only 5. The large disproportion in the data is additionally demonstrated by the coefficient of variation amounting to over 260%. This proves that the analyzed data are very diverse. Therefore, there are two conclusions that emerge from this analysis. First of all, a significant proportion of the meetings did not have any effect in the form of cooperation. Secondly, there are large disproportions in the values determining the number of partnerships undertaken for the analyzed data set.
As already mentioned, the indicated large differences in the sets, concerning the number of meetings and the number of partnerships, could be related to the existence of factors that significantly differentiate the analyzed events in the studied aspects.
Therefore, in this study, two categories of factors were selected, the impact of which on the numbers of meetings and partnerships was examined. The first factor is a geographical one, related to the place where the event took place, i.e., in which part of Europe and in which country. The second group of factors is related to the type of event, i.e., its form, and the type of partner supporting the event. The choice of such factors is related to the fact that the event organizer can decide on the event’s place, its form, and potential partners.
In order to analyze the impact of selected factors on the number of meetings and the number of partnerships during the analyzed events, limit values were defined for three intervals to determine the size level based on the 33rd and 66th percentiles. To determine them, only the cases were used in which at least one meeting took place. Thus, four classes were created to define the level of interest in the meetings, expressed as the number of meetings and the number of partnerships.
In the case of the number of meetings, the division into individual classes based on the above-mentioned percentiles was as follows:
Level 0—no meetings;
Level 1—low number of meetings—for events with 1 to 61 meetings;
Level 2—average number of meetings—for events with 62 to 142 meetings;
Level 3—large number of meetings—for events with more than 142 meetings.
For the number of partnerships, individual promotional events were grouped according to the following scheme:
Level 0—no partnerships;
Level 1—low number of partnerships—for events with 1 to 9 partnerships;
Level 2—average number of partnerships—for events with 10 to 32 partnerships;
Level 3—large number of partnerships—for events with more than 32 partnerships.
Moreover, in the analysis of the influence of selected factors, the additional basic measures of location and dispersion were compared for the number of meetings and for the number of instances of cooperation, respectively. In order to verify the statistical significance of the indicated differences, appropriate statistical tests were used.
In order to verify the occurrence of differences in the structure of the number of meetings and the number of partnerships, the chi-square test of independence, also called the Pearson independence test [
47], was used. The comparative analysis of the position measures was performed using the one-way ANOVA test and Student’s
t-test [
47], which allows determining the significance of the variation in the arithmetic mean among the indicated groups. Since these tests require meeting relevant assumptions [
47], in situations where applying them was not possible (the data did not meet the appropriate assumptions), one of the two following tests was used—the Mann–Whitney U test and the Kruskal–Wallis test [
48,
49]—in which, instead of the arithmetic mean values, the significance of the differentiation of a given characteristic is determined by comparing the medians. These tests require data to meet far fewer assumptions. In the analyses in question, all the assumptions required for the described tests were met. The hypotheses were verified with a 5% significance level. The above-described tests and analyses were carried out with the Statgraphics 18 software (Statgraphics Technologies, Inc., The Plains, VA, USA), while the graphs and drawings included in the article were prepared using Excel (Microsoft, Redmond, WA, USA) and Corel Draw software (Corel, Ottawa, Canada).