User:Staeiou/Journal/SDPatrolBot
< User:Staeiou | Journal
See en:User:SDPatrolBot/AltWarningResults
Edits after message
[edit]> summary(glm(formula = (edits_after_msg + deleted_edits_after_msg) ~ is_alt * (all_ns0_edits_30b + all_ns1_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), data = sdbot_jan15))
Call:
glm(formula = (edits_after_msg + deleted_edits_after_msg) ~ is_alt *
(all_ns0_edits_30b + all_ns1_edits_30b + ns3_edits_30days_before_msg +
seconds_between_reg_warn), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-280.28 -7.58 -4.43 0.59 1251.84
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 6.654e+00 3.177e+00 2.094 0.03653 *
is_alt -7.245e+00 4.590e+00 -1.578 0.11487
all_ns0_edits_30b 3.856e-01 1.273e-01 3.028 0.00254 **
all_ns1_edits_30b 1.158e+00 2.256e+00 0.513 0.60785
ns3_edits_30days_before_msg 5.163e-01 1.925e+00 0.268 0.78859
seconds_between_reg_warn 3.472e-10 7.134e-08 0.005 0.99612
is_alt:all_ns0_edits_30b 1.579e+00 1.816e-01 8.695 < 2e-16 ***
is_alt:all_ns1_edits_30b -2.053e+00 3.407e+00 -0.603 0.54695
is_alt:ns3_edits_30days_before_msg 2.082e-01 3.006e+00 0.069 0.94480
is_alt:seconds_between_reg_warn -1.307e-10 1.059e-07 -0.001 0.99902
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for gaussian family taken to be 3102)
Null deviance: 3351590 on 826 degrees of freedom
Residual deviance: 2534334 on 817 degrees of freedom
AIC: 9007.8
Number of Fisher Scoring iterations: 2
Article edits after
[edit]- Independently, users with more ns0 edits before message make more future ns0 edits after
- For editors with fewer than 7 ns0 edits before message, number of future ns0 edits is less with alt than default
- For editors with more than 7 ns0 edits before message, number of future ns0 edits is more with alt than default
summary(glm(formula = all_ns0_edits_30a ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), data = sdbot_jan15))
Call:
glm(formula = all_ns0_edits_30a ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-175.383 -5.033 -2.509 1.143 257.341
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 4.336e+00 1.425e+00 3.044 0.00241 **
is_alt -5.558e+00 2.058e+00 -2.700 0.00707 **
all_ns1_edits_30b 1.934e-01 1.012e+00 0.191 0.84848
all_ns0_edits_30b 3.485e-01 5.711e-02 6.102 1.61e-09 ***
ns3_edits_30days_before_msg -4.282e-01 8.631e-01 -0.496 0.61992
seconds_between_reg_warn -3.325e-09 3.199e-08 -0.104 0.91725
is_alt:all_ns1_edits_30b -1.602e+00 1.528e+00 -1.048 0.29480
is_alt:all_ns0_edits_30b 8.949e-01 8.144e-02 10.989 < 2e-16 ***
is_alt:ns3_edits_30days_before_msg -9.739e-02 1.348e+00 -0.072 0.94242
is_alt:seconds_between_reg_warn 2.998e-08 4.751e-08 0.631 0.52824
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for gaussian family taken to be 623.7657)
Null deviance: 834647 on 826 degrees of freedom
Residual deviance: 509617 on 817 degrees of freedom
AIC: 7681.2
Number of Fisher Scoring iterations: 2
Article talk edits
[edit]- Previous ns1 and ns3 edits before message are independently the strongest predictors of future ns1 edits
- Editors with more ns0 edits before message make more ns1 edits after receiving the alt message
- Editors with more ns1 and ns3 edits before message make fewer ns1 edits after receiving the alt message
> summary(glm(formula = all_ns1_edits_30a ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), data = sdbot_jan15))
Call:
glm(formula = all_ns1_edits_30a ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-8.413 -0.705 -0.516 0.471 31.846
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 5.282e-01 1.114e-01 4.743 2.49e-06 ***
is_alt 9.550e-03 1.609e-01 0.059 0.95269
all_ns1_edits_30b 4.823e-01 7.911e-02 6.097 1.67e-09 ***
all_ns0_edits_30b -6.107e-03 4.465e-03 -1.368 0.17176
ns3_edits_30days_before_msg 1.937e-01 6.748e-02 2.870 0.00421 **
seconds_between_reg_warn 1.721e-09 2.501e-09 0.688 0.49153
is_alt:all_ns1_edits_30b -2.800e-01 1.195e-01 -2.344 0.01933 *
is_alt:all_ns0_edits_30b 4.786e-02 6.367e-03 7.517 1.48e-13 ***
is_alt:ns3_edits_30days_before_msg -2.972e-01 1.054e-01 -2.820 0.00492 **
is_alt:seconds_between_reg_warn -2.232e-09 3.714e-09 -0.601 0.54808
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for gaussian family taken to be 3.812951)
Null deviance: 4798.7 on 826 degrees of freedom
Residual deviance: 3115.2 on 817 degrees of freedom
AIC: 3465.7
Number of Fisher Scoring iterations: 2
User talk edits
[edit]- Previous ns0 and ns3 edits are both independently predictors of future ns3 edits
- Editors with either more ns0 or ns3 edits before receiving the message make more ns3 edits with the alt message
> summary(glm(formula = ns3_edits_30days_after_msg ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), data = sdbot_jan15))
Call:
glm(formula = ns3_edits_30days_after_msg ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-9.1072 -0.5086 -0.4328 -0.3522 27.3501
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 4.043e-01 1.440e-01 2.807 0.005119 **
is_alt -1.308e-01 2.081e-01 -0.629 0.529765
all_ns1_edits_30b 1.689e-02 1.023e-01 0.165 0.868874
all_ns0_edits_30b 1.427e-02 5.773e-03 2.472 0.013636 *
ns3_edits_30days_before_msg 2.279e-01 8.725e-02 2.613 0.009152 **
seconds_between_reg_warn -1.066e-09 3.234e-09 -0.330 0.741704
is_alt:all_ns1_edits_30b -6.289e-02 1.545e-01 -0.407 0.684025
is_alt:all_ns0_edits_30b 2.510e-02 8.233e-03 3.048 0.002376 **
is_alt:ns3_edits_30days_before_msg 4.733e-01 1.363e-01 3.473 0.000542 ***
is_alt:seconds_between_reg_warn 4.772e-09 4.803e-09 0.994 0.320731
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for gaussian family taken to be 6.374625)
Null deviance: 6352.5 on 826 degrees of freedom
Residual deviance: 5208.1 on 817 degrees of freedom
AIC: 3890.7
Number of Fisher Scoring iterations: 2
Future SDPatrolBot warnings
[edit]- Editors with more ns0 edits before receiving the alt message are more likely to get warned again, compared to the default.
> summary(glm(formula = sdbot_warnings > 1 ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = sdbot_warnings > 1 ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-1.7326 -0.8508 -0.8072 1.4329 2.0028
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -9.337e-01 1.293e-01 -7.220 5.18e-13 ***
is_alt 4.998e-02 1.850e-01 0.270 0.7871
all_ns1_edits_30b -2.351e-02 9.224e-02 -0.255 0.7988
all_ns0_edits_30b -6.801e-03 7.348e-03 -0.925 0.3547
ns3_edits_30days_before_msg 4.356e-02 7.881e-02 0.553 0.5805
seconds_between_reg_warn 2.831e-09 2.689e-09 1.053 0.2924
is_alt:all_ns1_edits_30b 1.276e-01 1.331e-01 0.959 0.3377
is_alt:all_ns0_edits_30b 2.476e-02 1.065e-02 2.324 0.0201 *
is_alt:ns3_edits_30days_before_msg -1.715e-01 1.315e-01 -1.304 0.1923
is_alt:seconds_between_reg_warn 4.601e-09 3.975e-09 1.157 0.2471
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 1032.8 on 826 degrees of freedom
Residual deviance: 1006.3 on 817 degrees of freedom
AIC: 1026.3
Number of Fisher Scoring iterations: 4
Reported at AIV
[edit]- No statistically significant effects.
- Very marginal (p = 0.0942) interaction such that editors with more prior ns1 edits are more likely to be warned.
> summary(glm(formula = aiv_report ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = aiv_report ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b +
ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-0.8822 -0.3731 -0.3407 -0.3066 2.5230
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -2.589e+00 2.395e-01 -10.811 <2e-16 ***
is_alt -4.855e-01 3.554e-01 -1.366 0.1719
all_ns1_edits_30b -1.193e-01 1.670e-01 -0.714 0.4751
all_ns0_edits_30b -1.363e-02 2.058e-02 -0.662 0.5077
ns3_edits_30days_before_msg 1.572e-01 1.216e-01 1.293 0.1959
seconds_between_reg_warn -2.880e-09 6.450e-09 -0.446 0.6552
is_alt:all_ns1_edits_30b 3.419e-01 2.043e-01 1.674 0.0942 .
is_alt:all_ns0_edits_30b 2.703e-02 2.141e-02 1.262 0.2070
is_alt:ns3_edits_30days_before_msg -2.505e-01 2.430e-01 -1.031 0.3026
is_alt:seconds_between_reg_warn 7.576e-09 7.992e-09 0.948 0.3432
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 382.96 on 826 degrees of freedom
Residual deviance: 371.65 on 817 degrees of freedom
AIC: 391.65
Number of Fisher Scoring iterations: 6
Blocked
[edit]- Marginal effects: the more ns0 and ns3 edits before receiving the message, the more likely an editor is to be blocked
- The longer a user has been registered, the less likely they are to be blocked
- No effects from message sent
> summary(glm(formula = num_blocks > 0 ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = num_blocks > 0 ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-1.5964 -0.7562 -0.7390 -0.3374 2.5265
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -1.180e+00 1.408e-01 -8.380 <2e-16 ***
is_alt 6.773e-02 1.969e-01 0.344 0.7308
all_ns1_edits_30b -1.098e-01 1.236e-01 -0.888 0.3743
all_ns0_edits_30b 1.082e-02 5.808e-03 1.862 0.0626 .
ns3_edits_30days_before_msg 1.612e-01 9.349e-02 1.724 0.0846 .
seconds_between_reg_warn -1.178e-08 5.333e-09 -2.208 0.0272 *
is_alt:all_ns1_edits_30b 1.552e-01 1.604e-01 0.967 0.3333
is_alt:all_ns0_edits_30b -8.552e-03 7.618e-03 -1.122 0.2617
is_alt:ns3_edits_30days_before_msg -1.303e-01 1.297e-01 -1.004 0.3153
is_alt:seconds_between_reg_warn 4.376e-09 6.823e-09 0.641 0.5213
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 905.73 on 826 degrees of freedom
Residual deviance: 880.56 on 817 degrees of freedom
AIC: 900.56
Number of Fisher Scoring iterations: 5
Blocked
[edit]- Editors who registered earlier are less likely to be blocked
- Editors with more ns3 edits after are more likely to be blocked
- But for editors who made more ns3 edits after, those who received the alt message are less likely to be blocked
- Editors with more ns1 edits after are less likely to be blocked
- But for editors who made more ns1 edits after, those who received the alt message are more likely to be blocked
> summary(glm(formula = num_blocks > 0 ~ is_alt * (all_ns0_edits_30b + all_ns1_edits_30b + ns3_edits_30days_before_msg + all_ns0_edits_30a + all_ns1_edits_30a + ns3_edits_30days_after_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = num_blocks > 0 ~ is_alt * (all_ns0_edits_30b +
all_ns1_edits_30b + ns3_edits_30days_before_msg + all_ns0_edits_30a +
all_ns1_edits_30a + ns3_edits_30days_after_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-2.1141 -0.7542 -0.7016 -0.2993 2.2701
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -1.169e+00 1.440e-01 -8.120 4.65e-16 ***
is_alt 7.461e-02 2.026e-01 0.368 0.712639
all_ns0_edits_30b 8.392e-03 6.202e-03 1.353 0.176057
all_ns1_edits_30b -2.607e-02 1.068e-01 -0.244 0.807070
ns3_edits_30days_before_msg 1.577e-01 9.034e-02 1.746 0.080782 .
all_ns0_edits_30a 9.863e-04 6.911e-03 0.143 0.886513
all_ns1_edits_30a -3.095e-01 1.122e-01 -2.759 0.005802 **
ns3_edits_30days_after_msg 2.798e-01 7.901e-02 3.541 0.000399 ***
seconds_between_reg_warn -1.104e-08 5.415e-09 -2.039 0.041420 *
is_alt:all_ns0_edits_30b -1.060e-02 9.777e-03 -1.084 0.278258
is_alt:all_ns1_edits_30b 8.802e-02 1.493e-01 0.590 0.555435
is_alt:ns3_edits_30days_before_msg -1.606e-01 1.357e-01 -1.184 0.236314
is_alt:all_ns0_edits_30a 2.536e-03 8.445e-03 0.300 0.763979
is_alt:all_ns1_edits_30a 2.633e-01 1.303e-01 2.021 0.043245 *
is_alt:ns3_edits_30days_after_msg -2.390e-01 9.281e-02 -2.575 0.010029 *
is_alt:seconds_between_reg_warn 3.390e-09 6.880e-09 0.493 0.622170
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 905.73 on 826 degrees of freedom
Residual deviance: 859.30 on 811 degrees of freedom
AIC: 891.3
Number of Fisher Scoring iterations: 5
Long-term retention (edits 30-60 days after)
[edit]- Marginally significant: the longer a user has been registered and the more ns3 edits made before predict future retention
- No effects from message sent
> summary(glm(formula = edits_after_msg_30_60_days > 0 ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = edits_after_msg_30_60_days > 0 ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-0.9124 -0.2313 -0.2156 -0.2118 3.1707
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -3.794e+00 4.103e-01 -9.247 <2e-16 ***
is_alt -2.886e-02 5.509e-01 -0.052 0.9582
all_ns1_edits_30b -1.646e+00 1.059e+00 -1.555 0.1200
all_ns0_edits_30b 1.092e-02 9.154e-03 1.193 0.2329
ns3_edits_30days_before_msg 3.090e-01 1.613e-01 1.916 0.0554 .
seconds_between_reg_warn 1.078e-08 6.009e-09 1.795 0.0727 .
is_alt:all_ns1_edits_30b 1.687e+00 1.081e+00 1.561 0.1186
is_alt:all_ns0_edits_30b 7.188e-03 1.164e-02 0.617 0.5369
is_alt:ns3_edits_30days_before_msg -8.160e-02 1.963e-01 -0.416 0.6777
is_alt:seconds_between_reg_warn -1.993e-09 8.038e-09 -0.248 0.8042
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 231.08 on 826 degrees of freedom
Residual deviance: 207.15 on 817 degrees of freedom
AIC: 227.15
Number of Fisher Scoring iterations: 9
Warning message:
glm.fit: fitted probabilities numerically 0 or 1 occurred
Article deleted
[edit]- Marginally significant: for editors with more ns3 edits before message, less likely to have article deleted
- No effects from message sent
> summary(glm(formula = num_deletions > 0 ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = num_deletions > 0 ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-2.2704 0.5027 0.5714 0.5751 1.3139
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.728e+00 1.573e-01 10.985 <2e-16 ***
is_alt 1.243e-02 2.319e-01 0.054 0.9573
all_ns1_edits_30b 1.532e-01 1.078e-01 1.421 0.1553
all_ns0_edits_30b -2.445e-03 5.535e-03 -0.442 0.6587
ns3_edits_30days_before_msg -1.667e-01 8.785e-02 -1.897 0.0578 .
seconds_between_reg_warn 7.399e-10 3.829e-09 0.193 0.8468
is_alt:all_ns1_edits_30b 3.969e-02 2.206e-01 0.180 0.8572
is_alt:all_ns0_edits_30b -6.256e-03 7.725e-03 -0.810 0.4180
is_alt:ns3_edits_30days_before_msg 1.480e-01 1.437e-01 1.030 0.3030
is_alt:seconds_between_reg_warn 4.155e-09 6.277e-09 0.662 0.5080
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 692.01 on 826 degrees of freedom
Residual deviance: 680.71 on 817 degrees of freedom
AIC: 700.71
Number of Fisher Scoring iterations: 5
Article currently deleted
[edit]- Marginally significant: for editors with more ns0 edits after, less likely to have article deleted
- No effects from message sent
> summary(glm(formula = article_currently_deleted ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), family = binomial(link = "logit"), data = sdbot_jan15))
Call:
glm(formula = article_currently_deleted ~ is_alt * (all_ns1_edits_30b +
all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn),
family = binomial(link = "logit"), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-2.3631 0.3603 0.4083 0.4237 1.5794
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 2.415e+00 1.986e-01 12.163 <2e-16 ***
is_alt 3.481e-01 3.043e-01 1.144 0.2527
all_ns1_edits_30b 9.587e-02 1.257e-01 0.763 0.4455
all_ns0_edits_30b -9.778e-03 5.061e-03 -1.932 0.0534 .
ns3_edits_30days_before_msg -9.908e-02 1.025e-01 -0.966 0.3338
seconds_between_reg_warn -1.061e-09 4.381e-09 -0.242 0.8086
is_alt:all_ns1_edits_30b -2.566e-01 1.707e-01 -1.503 0.1327
is_alt:all_ns0_edits_30b -7.521e-03 8.509e-03 -0.884 0.3768
is_alt:ns3_edits_30days_before_msg -3.082e-03 1.463e-01 -0.021 0.9832
is_alt:seconds_between_reg_warn -4.213e-09 5.981e-09 -0.704 0.4812
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 503.03 on 826 degrees of freedom
Residual deviance: 480.60 on 817 degrees of freedom
AIC: 500.6
Number of Fisher Scoring iterations: 5
Proportion of future ns0 edits deleted
[edit]- Editors with more ns0 edits before have a lower proportion of future ns0 edits deleted
> summary(glm(formula = deleted_ns0_edits_30days_after_msg/all_ns0_edits_30a ~ is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg + seconds_between_reg_warn), data = sdbot_jan15))
Call:
glm(formula = deleted_ns0_edits_30days_after_msg/all_ns0_edits_30a ~
is_alt * (all_ns1_edits_30b + all_ns0_edits_30b + ns3_edits_30days_before_msg +
seconds_between_reg_warn), data = sdbot_jan15)
Deviance Residuals:
Min 1Q Median 3Q Max
-0.8359 -0.1639 0.1973 0.2097 0.6361
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 8.030e-01 2.652e-02 30.281 < 2e-16 ***
is_alt 1.293e-02 3.825e-02 0.338 0.735426
all_ns1_edits_30b 5.058e-03 1.748e-02 0.289 0.772474
all_ns0_edits_30b -3.169e-03 8.614e-04 -3.679 0.000257 ***
ns3_edits_30days_before_msg -2.527e-03 1.471e-02 -0.172 0.863675
seconds_between_reg_warn 2.505e-10 5.794e-10 0.432 0.665635
is_alt:all_ns1_edits_30b -1.548e-02 2.526e-02 -0.613 0.540188
is_alt:all_ns0_edits_30b -7.415e-04 1.227e-03 -0.604 0.546026
is_alt:ns3_edits_30days_before_msg -3.034e-02 2.144e-02 -1.416 0.157488
is_alt:seconds_between_reg_warn -4.833e-10 8.112e-10 -0.596 0.551580
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for gaussian family taken to be 0.1365316)
Null deviance: 81.385 on 556 degrees of freedom
Residual deviance: 74.683 on 547 degrees of freedom
(270 observations deleted due to missingness)
AIC: 483.51
Number of Fisher Scoring iterations: 2
>