FINDINGS: books for younger kids have fewer edges, but effect goes away when you control for author in mixed effect model….but missing some data.
If you limit to books since 1975, potentially intersteing interactions between year and age.
##
## Call:
## lm(formula = mean_edges ~ earliest_advert_level_months, data = stats_by_book)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.039581 -0.010410 -0.000905 0.009195 0.055213
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.667e-02 2.880e-03 16.203 <2e-16 ***
## earliest_advert_level_months 1.655e-04 7.332e-05 2.257 0.0249 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.01623 on 236 degrees of freedom
## (6 observations deleted due to missingness)
## Multiple R-squared: 0.02113, Adjusted R-squared: 0.01698
## F-statistic: 5.094 on 1 and 236 DF, p-value: 0.02492
## Linear mixed model fit by REML ['lmerMod']
## Formula: mean_edges ~ earliest_advert_level_months + (1 | author)
## Data: stats_by_book
##
## REML criterion at convergence: -1278
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.35745 -0.46372 -0.00944 0.39628 3.06005
##
## Random effects:
## Groups Name Variance Std.Dev.
## author (Intercept) 0.0001490 0.01221
## Residual 0.0001276 0.01130
## Number of obs: 238, groups: author, 175
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 4.877e-02 3.034e-03 16.074
## earliest_advert_level_months 1.198e-04 7.361e-05 1.628
##
## Correlation of Fixed Effects:
## (Intr)
## erlst_dvr__ -0.916
## Linear mixed model fit by REML ['lmerMod']
## Formula: mean_edges ~ earliest_advert_level_months * year + (1 | author)
## Data: stats_by_book %>% filter(year > 1975)
##
## REML criterion at convergence: -1097.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.26921 -0.45143 -0.02777 0.40766 3.09373
##
## Random effects:
## Groups Name Variance Std.Dev.
## author (Intercept) 0.0001266 0.01125
## Residual 0.0001422 0.01192
## Number of obs: 212, groups: author, 159
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) -4.109e-01 5.539e-01 -0.742
## earliest_advert_level_months 2.275e-02 1.405e-02 1.619
## year 2.296e-04 2.762e-04 0.831
## earliest_advert_level_months:year -1.129e-05 7.006e-06 -1.611
##
## Correlation of Fixed Effects:
## (Intr) erl___ year
## erlst_dvr__ -0.927
## year -1.000 0.927
## erlst_dv__: 0.927 -1.000 -0.927
## fit warnings:
## Some predictor variables are on very different scales: consider rescaling