Tables 4 and 5 report the GLS regression results for the diversity of market and R&D moves in the established Sonite and new Vodite markets, respectively. The results support our main findings in Tables 2 and 3. These results also suggest that when the overall industry shrinks, firms are especially likely to use diversifying moves into the new Vodite market to identify new opportunities. Together, these analyses provide strong confirmation of our main propositions.
Additional analyses
We conducted additional analyses. First, we further probed the relationship between firm performance and competitive moves. While our evolutionary search arguments suggested and found a continuous relationship between performance and competitive moves, we also considered whether firms may initiate search when confronted with failure relative to a reference point (Cyert and March, 1963; Greve, 2003b). Based on data from the Markstrat environment and participant interviews indicating that competitors used their rivals' performance as their benchmark (this relative performance also influenced participants' class grades), we used industry mean as a reference point—i.e., we expected that performance below the mean compared to the rest of the teams in the industry would lead the focal firm to alter the number and diversity of its competitive moves. Following Greve (2003b), we used the equation Yr+1 = f[β1(Pr − Lr)IPr>Lr + β2(Pr − Lr)IPr⩽Lr + β3Xr] where the outcome variable is Yr+1, (competitive move frequency and competitive move diversity), β1 is the effect of performance when performance is above the reference point, and β2 is the effect when performance is below the reference point. Pr signifies a firm's performance while Lr signifies the reference point (mean industry market share in the round). The indicator variable, I, takes the value of 1 if the subscript was true and 0 otherwise. Xr is a vector for control variables. This equation enabled us to test if performance relative to a reference point leads firms to alter the frequency (or diversity) of their competitive moves. Significance tests of the coefficients provide evidence of an effect on competitive moves, while the comparison of the coefficients indicates how the effect changes going from below to above a reference point.
Our results (details available from the authors) indicate that the slope of the performance variable indeed shifts, but the sign of the coefficients remains unchanged for below versus above the reference point. For example, in the established market, the coefficients for below-the-mean and above-the-mean performance variables are both negative and significant in predicting frequency of market moves and both positive and significant in predicting frequency of R&D moves. But there is a slight variation in the value of the coefficients. The negative relationship between performance and market moves is more pronounced for below-the-mean performers and, in contrast, the positive relationship between performance and R&D moves is more pronounced for above-the-mean performers. That is, a decline in performance is particularly likely to encourage those firms that perform worse than the upper half of the industry to use more market moves. Similarly, an improvement in performance is particularly likely to encourage those firms that perform better than the lower half of the industry to use more R&D moves. These findings are consistent with our hypothesized effects in established markets that emphasize the attractiveness of market versus R&D moves to low- versus high-performing firms, respectively. Thus, these results further support our original findings.
Уважаемый посетитель!
Чтобы распечатать файл, скачайте его (в формате Word).
Ссылка на скачивание - внизу страницы.