Regression Models to Explain Double Jeopardy in Long Tail with Log Average Time Spent as the Dependent Variable
Statistic | Model 1 | Model 2 | Model 3 | Model 4 | ||||
---|---|---|---|---|---|---|---|---|
β | T | β | T | β | T | β | T | |
Unique (log) | .262** | 33.07 | .264** | 33.27 | .266** | 33.49 | .083** | 14.23 |
News | −.084** | −9.82 | −.084** | −9.87 | −.115** | −18.60 | ||
Lifestyle | −.125** | −11.28 | −.125** | −11.28 | −.084** | −10.43 | ||
Sports | .167** | 8.61 | .166** | 8.57 | −.035** | −2.47 | ||
Months | † | † | † | † | ||||
Visits (root) | .536** | 147.14 | ||||||
Control | ||||||||
Intercept | −.052* | −1.98 | −.040 | −1.51 | −.058 | −2.05 | −.309* | −14.83 |
Adjusted R2 | .043 | .055 | .056 | .503 |
Note. N = 24,000. The dependent variable is the logarithm of average total minutes in millions.
↵† After adding monthly data as an independent variable, the author found that there were no significant effects on beta values.
*p < .05. **p < .01.