Table 2 presents the descriptive statistics—in addition to the above eight variables, physical proximity; teamwork, customer, and presence from Koren & Pető;6 and remote working from Dingel & Neiman4 are listed with their values and correlations as per the author’s calculation. In general, 0.7 or higher is considered to be an acceptable reliability coefficient,12 but several researchers have accepted 0.6 or more.13,14 According to the latter criterion, all eight variables were acceptable, but according to the former, mechanical movement and horizontal teamwork were unacceptable. Therefore, we prepared a three-item version without work context-time pressure for mechanical movement, and a three-item version with work context-coordinate or lead others for horizontal teamwork, which had been dropped in the first factor analysis for being the same as the other two items but loaded on other factors. After confirming that the reliability coefficients of these variables were 0.7 or higher, they were used for the following analysis alternatively (the result is omitted because it did not make a big difference).
Looking at the magnitude of the correlation coefficient, all except leadership and mechanical movement showed a statistically significant correlation with physical proximity at the 5% level. However, values exceeding 0.1, which is the standard value for “small” by Cohen (1988), were observed in three variables: response to aggression (r = 0.456, p < 0.01), autonomy (r=-0.130, p < 0.05), and horizontal teamwork (r = 0.306, p < 0.01). Of these, only response to aggression and horizontal teamwork exceeded the standard value of 0.3 for “medium” by Cohen (1988).15 Moreover, as shown in the table, among the three from Koren & Pető,6 customer (r = 0.415, p < 0.01) and presence (r = 0.197, p < 0.01) exceeded 0.1, with only customer exceeding 0.3. Dingel & Neiman’s remote working4 (r=-0.356, p < 0.01) also exceeded 0.3. However, the coefficient of Koren & Pető’s teamwork6 (r = 0.098, p < 0.01) was below 0.1. This is probably because, as described at the beginning, items that represent horizontal teamwork and those that represent vertical teamwork are mixed in the variable.
Table 3 shows the results of regression analysis. Eight independent variables were individually inputted into the regression equation. Response to aggression (β = 0.456, p < 0.01) and horizontal teamwork (β = 0.306, p < 0.01) showed a significant positive correlation, supporting H1. Information processing (β=-0.085, p < 0.01), autonomy (β=-0.129, p < 0.01), and communication with the outside (β=-0.065, p < 0.05) showed a significant negative correlation, supporting H2. Furthermore, a significant positive correlation was observed in adverse conditions (β = 0.074, p < 0.05), supporting H3. However, regarding leadership (β = 0.057, p > 0.05), where a positive correlation was expected, and mechanical movement (β = 0.001, p > 0.05), where a negative correlation was expected, a significant correlation was not found at the 5% level. Comparing the adjusted R-squared value, response to aggression was the largest at 0.207, followed by horizontal teamwork at 0.093 and autonomy at 0.016. This shows that response to aggression has the greatest influence on social distancing, and it can thus be said that this is the result that supports H4.
Table 2
| Mean | SD | α | Physical Proximity | Adverse Conditions | Leadership | Information Processing | Response to Aggression | Mechanical Movement | Autonomy | Communication with the Outside | Horizontal Teamwork | aTeamwork | aCustomer | aPresence |
Physical Proximity | 60.300 | 16.875 | | | | | | | | | | | | | |
Adverse Conditions | 18.542 | 17.538 | 0.929 | 0.074* | | | | | | | | | | | |
Leadership | 48.440 | 11.806 | 0.939 | 0.056 | -0.058 | | | | | | | | | | |
Information Processing | 66.133 | 11.940 | 0.923 | -0.085** | -0.307** | 0.553** | | | | | | | | | |
Response to Aggression | 36.748 | 12.878 | 0.871 | 0.456** | -0.034 | 0.243** | 0.146** | | | | | | | | |
Mechanical Movement | 58.955 | 9.699 | 0.694 | 0.001 | -0.038 | -0.083** | 0.195** | 0.186** | | | | | | | |
Autonomy | 76.918 | 11.440 | 0.897 | -0.130** | -0.214** | 0.349** | 0.417** | 0.041 | -0.137** | | | | | | |
Communication with the Outside | 67.569 | 18.633 | 0.810 | -0.064* | -0.314** | 0.447** | 0.585** | 0.312** | 0.084** | 0.534** | | | | | |
Horizontal Teamwork | 84.353 | 8.904 | 0.656 | 0.306** | -0.052 | 0.438** | 0.392** | 0.373** | 0.188** | 0.155** | 0.385** | | | | |
aTeamwork | 55.408 | 12.084 | 0.889 | 0.098** | -0.094** | 0.933** | 0.588** | 0.287** | -0.024 | 0.288** | 0.485** | 0.619** | | | |
aCustomer | 53.837 | 14.099 | 0.775 | 0.415** | -0.312** | 0.525** | 0.412** | 0.584** | -0.068* | 0.363** | 0.626** | 0.368** | 0.526** | | |
aPresence | 38.202 | 19.354 | 0.911 | 0.197** | 0.814** | -0.040 | -0.258** | -0.064* | 0.082* | -0.272** | -0.403** | -0.060 | -0.107** | -0.259** | |
bRemote Working | 57.767 | 13.832 | 0.914 | -0.356** | -0.838** | -0.067* | 0.176** | -0.147** | -0.042 | 0.177** | 0.255** | -0.061 | 0.011 | -0.029 | 0.920** |
Note(s): n = 968; *Significance at the 5% level; **Significance at the 1% level; a is from Koren & Pető,6 b is from Dingel & Neiman.4 |
In addition, as shown in the table, all three from Koren & Pető6—teamwork (β = 0.098, p < 0.01), customer (β = 0.414, p < 0.01), and presence (β = 0.198, p < 0.01)—also showed a significant correlation at the 1% level. However, among them, the adjusted R-squared value of customer was significantly large at 0.171. Dingel & Neiman4’s remote working (β=-0.356, p < 0.01) also showed a significant correlation at the 1% level, and its adjusted R-squared value was relatively large at 0.126. Comparing the magnitude of the partial regression coefficient and the adjusted R-squared value, the order is as follows: response to aggression (β = 0.456, R2 = 0.207), customer (β = 0.414, R2 = 0.171), and remote working (β=-0.356, R2 = 0.126).
Table 3
Results of simple regression analysis with physical proximity as the dependent variable
Variable | β | | R2 | Adj-R2 | F | |
Adverse Conditions | 0.074 | * | 0.006 | 0.005 | 5.367 | * |
Leadership | 0.057 | | 0.003 | 0.002 | 3.094 | |
Information Processing | -0.085 | ** | 0.007 | 0.006 | 7.011 | ** |
Response to Aggression | 0.456 | ** | 0.208 | 0.207 | 253.458 | ** |
Mechanical Movement | 0.001 | | 0.000 | -0.001 | 0.001 | |
Autonomy | -0.129 | ** | 0.017 | 0.016 | 16.369 | ** |
Communication with the Outside | -0.065 | * | 0.004 | 0.003 | 4.056 | * |
Horizontal Teamwork | 0.306 | ** | 0.094 | 0.093 | 99.691 | ** |
aTeamwork | 0.098 | ** | 0.010 | 0.009 | 9.348 | ** |
aCustomer | 0.414 | ** | 0.172 | 0.171 | 200.134 | ** |
aPresence | 0.198 | ** | 0.039 | 0.038 | 39.194 | ** |
bRemote Working | -0.356 | ** | 0.127 | 0.126 | 139.189 | ** |
Note(s): n = 968; *Significance at the 5% level; **Significance at the 1% level; a from Koren & Pető,6 b from Dingel & Neiman.4 |
The first column of Table 4 shows the results of multiple regression analysis using the stepwise method. Response to aggression (β = 0.460, p < 0.01) and horizontal teamwork (β = 0.288, p < 0.01) showed a positive correlation, and information processing (β=-0.096, p < 0.01), mechanical movement (β=-0.099, p < 0.01), and communication with the outside (β=-0.254, p < 0.01) showed a negative correlation. Among them, mechanical movement did not show a significant correlation by simple correlation, but showed a significant negative correlation by multiple regression after controlling the effects of other variables, which is consistent with the hypothesis. On the other hand, adverse conditions and autonomy showed positive and negative correlations in single regression, respectively, but did not show a significant correlation in multiple regression after controlling the effects of other variables. Leadership showed no significant correlation in either single or multiple regression.
Looking at the results of the three items from Koren & Pető6 in the second column of the table, customer (β = 0.595, p < 0.01) and presence (β = 0.333, p < 0.01) showed a positive correlation, the same as in simple correlation, but teamwork (β=-0.179, p < 0.01) showed a negative correlation. It is theoretically difficult to think that teamwork has a negative correlation with physical proximity, so it can be interpreted as a problem of multicollinearity. This indicates that some caution is required when using the three items from Koren & Pető.6 The 14 variables that compose the three from Koren & Pető6 were simply averaged and subtracted from 100 to create one variable Social distancing Index (α = 0.727, mean value = 50.753, standard deviation = 9.085) following the method of Crowley & Doran,4 and the relationship with physical proximity was examined. Although the problem of multicollinearity was solved, the adjusted R-squared value dropped significantly (β = 0.437, p < 0.01, R2 = 0.190).
Therefore, in the third column, the results of a regression analysis performed on a variable group in which customer and remote working, whose adjusted R-squared value for single regression exceeds 0.13—the “medium” level15— are added to the first column model as shown. In addition to customer and remote working, four of the five variables shown in the first column, except mechanical movement, are significant. However, the magnitude of the partial correlation coefficient is different between the first and third columns, and it seems that there is a problem of multicollinearity. This is understandable considering that many of the items that make up customer and remote working were excluded during the exploratory factor analysis due to the load on multiple factors. Despite this problem, the fact that the variables in the first column, including response to aggression, became significant even after controlling for the impact of customer and remote working indicates the robustness of the model in the first column. For reference, the adjusted R-squared values for all the three models in Table 4—0.319, 0.293, and 0.491—exceed 0.26, the “large” level.15 Among them, the value of the mixed model in the third column is the highest. Therefore, although it is worrisome that the distortion of the partial correlation coefficient impairs the accuracy of the influence of individual variables, the mixed model in the third column is considered to be useful for predicting the feasibility of social distancing more accurately by considering differences in customer contact and outdoor activities.
Table 4
Results of multiple regression analysis with physical proximity as the dependent variable
Variable | β |
Adverse Conditions | | | - | | - | |
Leadership | | | - | | - | |
Information Processing | -0.096 | ** | - | | -0.158 | ** |
Response to Aggression | 0.460 | ** | - | | 0.155 | ** |
Mechanical Movement | -0.099 | ** | - | | | |
Autonomy | | | - | | - | |
Communication with the Outside | -0.254 | ** | - | | -0.421 | ** |
Horizontal Teamwork | 0.288 | ** | - | | 0.250 | ** |
aTeamwork | - | | -0.179 | ** | - | |
aCustomer | - | | 0.595 | ** | 0.565 | ** |
aPresence | - | | 0.333 | ** | - | |
bRemote Working | | | | | -0.199 | ** |
R2 | 0.322 | | 0.295 | | 0.494 | |
Adjusted R2 | 0.319 | | 0.293 | | 0.491 | |
F | 91.288 | ** | 134.161 | ** | 155.435 | ** |
Note(s): n = 968; *Significance at the 5% level; **Significance at the 1% level; a from Koren & Pető,6 b from Dingel & Neiman;4 “-” indicates that it is not used in the regression model; blank cells indicate not selected in stepwise regression. |
Note
The lower left box is one standard deviation lower than the average value; the upper right box is one standard deviation higher than the average value.
Figure 1 shows the correlation between response to aggression and physical proximity—the lower left box indicates that the values of both are one standard deviation lower than the average value, while the upper right box indicates that they are one standard deviation higher than the average value. Appendices A1 and A2 include the occupations from the two boxes for each variable. A1 comprises the so-called “experts,” while in A2, there are elementary/middle school and special education teachers, therapists, technicians, nurses, restaurants’ and entertainment facilities’ staff, travel/postal service clerks, flight/transportation attendants, etc.