Skip to content
Permalink
Branch: master
Find file Copy path
v1 a530c5a Apr 6, 2019
1 contributor

Users who have contributed to this file

311 lines (294 sloc) 19.6 KB
Misleading Science Journalism v1 data analysis
Total number of respondents across conditions: 302
Number of 'Other' gender respondents: 1
Removing other gender respondents
New total number of respondents across conditions (after removal): 301
Coding likerts and categoricals....
pp_dem pp_gop pp_green pp_libertarian
0 1 0 0 0
1 0 0 0 0
2 1 0 0 0
3 1 0 0 0
4 0 1 0 0
Done coding likerts and categoricals.
Setting up new df for dailybeast
dailybeast describe control group
confidence_suicide accuracy_journalism gender funding \
count 99.000000 99.000000 99.000000 99.000000
mean 2.656566 2.969697 0.505051 0.545455
std 1.161997 1.024740 0.502519 0.500464
min 0.000000 0.000000 0.000000 0.000000
25% 2.000000 2.000000 0.000000 0.000000
50% 3.000000 3.000000 1.000000 1.000000
75% 3.000000 4.000000 1.000000 1.000000
max 5.000000 5.000000 1.000000 1.000000
rate_of_suicide is_dailybeast pp_gop pp_dem
count 99.000000 99.0 99.000000 99.000000
mean 0.545455 0.0 0.232323 0.383838
std 0.500464 0.0 0.424463 0.488794
min 0.000000 0.0 0.000000 0.000000
25% 0.000000 0.0 0.000000 0.000000
50% 1.000000 0.0 0.000000 0.000000
75% 1.000000 0.0 0.000000 1.000000
max 1.000000 0.0 1.000000 1.000000
dailybeast describe experiment group
confidence_suicide accuracy_journalism gender funding \
count 101.000000 101.000000 101.000000 101.000000
mean 3.792079 2.247525 0.485149 0.821782
std 1.194293 1.388567 0.502272 0.384605
min 0.000000 0.000000 0.000000 0.000000
25% 3.000000 1.000000 0.000000 1.000000
50% 4.000000 2.000000 0.000000 1.000000
75% 5.000000 3.000000 1.000000 1.000000
max 5.000000 5.000000 1.000000 1.000000
rate_of_suicide is_dailybeast pp_gop pp_dem
count 101.000000 101.0 101.000000 101.000000
mean 0.782178 1.0 0.178218 0.534653
std 0.414824 0.0 0.384605 0.501285
min 0.000000 1.0 0.000000 0.000000
25% 1.000000 1.0 0.000000 0.000000
50% 1.000000 1.0 0.000000 1.000000
75% 1.000000 1.0 0.000000 1.000000
max 1.000000 1.0 1.000000 1.000000
Num rows in dailybeast_df: 200
confidence_suicide accuracy_journalism gender funding rate_of_suicide \
0 4.0 4 0 1.0 1.0
1 3.0 3 0 0.0 1.0
2 4.0 4 0 1.0 1.0
3 2.0 3 1 1.0 1.0
4 2.0 2 1 1.0 1.0
is_dailybeast pp_gop pp_dem
0 0 0 1
1 0 0 0
2 0 0 1
3 0 0 1
4 0 1 0
Optimization terminated successfully.
Current function value: 0.603063
Iterations 5
Dailybeast logit (cov_type = 'HC3') targeting 'rate_of_suicide', trained on:
['gender', 'is_dailybeast', 'pp_gop', 'pp_dem']
Dailybeast Logit Results:
Logit Regression Results
==============================================================================
Dep. Variable: rate_of_suicide No. Observations: 200
Model: Logit Df Residuals: 196
Method: MLE Df Model: 3
Date: Sat, 30 Mar 2019 Pseudo R-squ.: 0.05426
Time: 22:38:05 Log-Likelihood: -120.61
converged: True LL-Null: -127.53
LLR p-value: 0.003131
=================================================================================
coef std err z P>|z| [0.025 0.975]
---------------------------------------------------------------------------------
gender -0.1307 0.267 -0.490 0.624 -0.654 0.392
is_dailybeast 1.1338 0.288 3.933 0.000 0.569 1.699
pp_gop 0.0447 0.370 0.121 0.904 -0.680 0.769
pp_dem 0.3954 0.291 1.360 0.174 -0.174 0.965
=================================================================================
Dailybeast logit (cov_type = 'HC3') targeting 'funding':
Optimization terminated successfully.
Current function value: 0.571273
Iterations 5
Logit Regression Results
==============================================================================
Dep. Variable: funding No. Observations: 200
Model: Logit Df Residuals: 196
Method: MLE Df Model: 3
Date: Sat, 30 Mar 2019 Pseudo R-squ.: 0.08309
Time: 22:38:05 Log-Likelihood: -114.25
converged: True LL-Null: -124.61
LLR p-value: 0.0001211
=================================================================================
coef std err z P>|z| [0.025 0.975]
---------------------------------------------------------------------------------
gender -0.0484 0.278 -0.174 0.862 -0.593 0.497
is_dailybeast 1.2747 0.307 4.158 0.000 0.674 1.876
pp_gop 0.0705 0.368 0.192 0.848 -0.651 0.792
pp_dem 0.5344 0.303 1.764 0.078 -0.059 1.128
=================================================================================
Dailybeast ols (cov_type = 'HC3') targeting 'accuracy_journalism':
OLS Regression Results
===============================================================================
Dep. Variable: accuracy_journalism R-squared: 0.138
Model: OLS Adj. R-squared: 0.111
Method: Least Squares F-statistic: 5.303
Date: Sat, 30 Mar 2019 Prob (F-statistic): 4.38e-05
Time: 22:38:05 Log-Likelihood: -316.54
No. Observations: 200 AIC: 647.1
Df Residuals: 193 BIC: 670.2
Df Model: 6
Covariance Type: HC3
========================================================================================================
coef std err z P>|z| [0.025 0.975]
--------------------------------------------------------------------------------------------------------
Intercept 2.8618 0.212 13.527 0.000 2.447 3.277
C(pp_dem)[T.1] 0.3744 0.248 1.510 0.131 -0.112 0.861
C(is_dailybeast)[T.1] -1.1326 0.293 -3.865 0.000 -1.707 -0.558
C(pp_gop)[T.1] -0.2778 0.257 -1.080 0.280 -0.782 0.226
C(gender)[T.1] 0.0568 0.175 0.325 0.745 -0.286 0.400
C(pp_dem)[T.1]:C(is_dailybeast)[T.1] 0.4060 0.381 1.067 0.286 -0.340 1.152
C(pp_gop)[T.1]:C(is_dailybeast)[T.1] 0.6900 0.554 1.246 0.213 -0.395 1.775
==============================================================================
Omnibus: 3.428 Durbin-Watson: 2.042
Prob(Omnibus): 0.180 Jarque-Bera (JB): 3.248
Skew: -0.250 Prob(JB): 0.197
Kurtosis: 2.627 Cond. No. 10.9
==============================================================================
Warnings:
[1] Standard Errors are heteroscedasticity robust (HC3)
Setting up new df for nyt
nyt describe control group
confidence_weight accuracy_journalism gender fda_reg \
count 99.000000 99.000000 99.000000 99.000000
mean 3.545455 2.969697 0.505051 0.292929
std 1.090523 1.024740 0.502519 0.457422
min 1.000000 0.000000 0.000000 0.000000
25% 3.000000 2.000000 0.000000 0.000000
50% 4.000000 3.000000 1.000000 0.000000
75% 4.000000 4.000000 1.000000 1.000000
max 5.000000 5.000000 1.000000 1.000000
weight is_nyt pp_gop pp_dem
count 99.000000 99.0 99.000000 99.000000
mean 0.191919 0.0 0.232323 0.383838
std 0.395814 0.0 0.424463 0.488794
min 0.000000 0.0 0.000000 0.000000
25% 0.000000 0.0 0.000000 0.000000
50% 0.000000 0.0 0.000000 0.000000
75% 0.000000 0.0 0.000000 1.000000
max 1.000000 0.0 1.000000 1.000000
nyt describe experiment group
confidence_weight accuracy_journalism gender fda_reg \
count 101.000000 101.000000 101.000000 101.000000
mean 3.673267 2.871287 0.326733 0.405941
std 1.059329 1.101484 0.471358 0.493522
min 0.000000 0.000000 0.000000 0.000000
25% 3.000000 2.000000 0.000000 0.000000
50% 4.000000 3.000000 0.000000 0.000000
75% 4.000000 4.000000 1.000000 1.000000
max 5.000000 5.000000 1.000000 1.000000
weight is_nyt pp_gop pp_dem
count 101.000000 101.0 101.000000 101.000000
mean 0.564356 1.0 0.346535 0.405941
std 0.498314 0.0 0.478239 0.493522
min 0.000000 1.0 0.000000 0.000000
25% 0.000000 1.0 0.000000 0.000000
50% 1.000000 1.0 0.000000 0.000000
75% 1.000000 1.0 1.000000 1.000000
max 1.000000 1.0 1.000000 1.000000
Num rows in nyt_df: 200
confidence_weight accuracy_journalism gender fda_reg weight is_nyt \
0 4.0 4 0 1.0 0.0 0
1 2.0 3 0 1.0 0.0 0
2 3.0 4 0 0.0 0.0 0
3 3.0 3 1 0.0 0.0 0
4 3.0 2 1 0.0 0.0 0
pp_gop pp_dem
0 0 1
1 0 0
2 0 1
3 0 1
4 1 0
Optimization terminated successfully.
Current function value: 0.631483
Iterations 5
nyt logit (cov_type = 'HC3') targeting 'weight', trained on:
['gender', 'is_nyt', 'pp_gop', 'pp_dem']
nyt Logit Results:
Logit Regression Results
==============================================================================
Dep. Variable: weight No. Observations: 200
Model: Logit Df Residuals: 196
Method: MLE Df Model: 3
Date: Sat, 30 Mar 2019 Pseudo R-squ.: 0.04906
Time: 22:38:05 Log-Likelihood: -126.30
converged: True LL-Null: -132.81
LLR p-value: 0.004567
==============================================================================
coef std err z P>|z| [0.025 0.975]
------------------------------------------------------------------------------
gender -0.8156 0.265 -3.078 0.002 -1.335 -0.296
is_nyt 0.9819 0.300 3.274 0.001 0.394 1.570
pp_gop -0.5654 0.345 -1.637 0.102 -1.242 0.112
pp_dem -0.6087 0.300 -2.030 0.042 -1.196 -0.021
==============================================================================
nyt logit (cov_type = 'HC3') targeting 'fda_reg':
Optimization terminated successfully.
Current function value: 0.641197
Iterations 5
Logit Regression Results
==============================================================================
Dep. Variable: fda_reg No. Observations: 200
Model: Logit Df Residuals: 196
Method: MLE Df Model: 3
Date: Sat, 30 Mar 2019 Pseudo R-squ.: 0.009653
Time: 22:38:05 Log-Likelihood: -128.24
converged: True LL-Null: -129.49
LLR p-value: 0.4753
==============================================================================
coef std err z P>|z| [0.025 0.975]
------------------------------------------------------------------------------
gender -0.5887 0.268 -2.194 0.028 -1.115 -0.063
is_nyt 0.0387 0.272 0.142 0.887 -0.495 0.572
pp_gop 0.0978 0.324 0.302 0.763 -0.538 0.733
pp_dem -0.7158 0.295 -2.425 0.015 -1.294 -0.137
==============================================================================
nyt ols (cov_type = 'HC3') targeting 'accuracy_journalism' (kitchen sink model):
OLS Regression Results
===============================================================================
Dep. Variable: accuracy_journalism R-squared: 0.064
Model: OLS Adj. R-squared: 0.030
Method: Least Squares F-statistic: 2.288
Date: Sat, 30 Mar 2019 Prob (F-statistic): 0.0292
Time: 22:38:05 Log-Likelihood: -288.79
No. Observations: 200 AIC: 593.6
Df Residuals: 192 BIC: 620.0
Df Model: 7
Covariance Type: HC3
=================================================================================
coef std err z P>|z| [0.025 0.975]
---------------------------------------------------------------------------------
Intercept 3.0492 0.225 13.536 0.000 2.608 3.491
pp_dem 0.3403 0.248 1.371 0.170 -0.146 0.827
is_nyt -0.5882 0.303 -1.944 0.052 -1.181 0.005
pp_dem:is_nyt 0.2278 0.364 0.626 0.531 -0.485 0.941
pp_gop -0.3245 0.257 -1.262 0.207 -0.828 0.179
pp_gop:is_nyt 0.7534 0.390 1.929 0.054 -0.012 1.519
gender -0.2668 0.209 -1.275 0.202 -0.677 0.143
gender:is_nyt 0.3617 0.314 1.152 0.249 -0.254 0.977
==============================================================================
Omnibus: 6.261 Durbin-Watson: 1.792
Prob(Omnibus): 0.044 Jarque-Bera (JB): 6.412
Skew: -0.435 Prob(JB): 0.0405
Kurtosis: 2.891 Cond. No. 11.1
==============================================================================
Warnings:
[1] Standard Errors are heteroscedasticity robust (HC3)
nyt ols (cov_type = 'HC3') targeting 'accuracy_journalism' (constrained model):
OLS Regression Results
===============================================================================
Dep. Variable: accuracy_journalism R-squared: 0.040
Model: OLS Adj. R-squared: 0.020
Method: Least Squares F-statistic: 2.126
Date: Sat, 30 Mar 2019 Prob (F-statistic): 0.0790
Time: 22:38:06 Log-Likelihood: -291.39
No. Observations: 200 AIC: 592.8
Df Residuals: 195 BIC: 609.3
Df Model: 4
Covariance Type: HC3
==============================================================================
coef std err z P>|z| [0.025 0.975]
------------------------------------------------------------------------------
Intercept 2.8185 0.168 16.742 0.000 2.488 3.148
pp_dem 0.4539 0.175 2.594 0.009 0.111 0.797
pp_gop 0.0819 0.197 0.415 0.678 -0.305 0.469
gender -0.0832 0.153 -0.543 0.587 -0.384 0.217
is_nyt -0.1326 0.150 -0.886 0.375 -0.426 0.161
==============================================================================
Omnibus: 4.390 Durbin-Watson: 1.777
Prob(Omnibus): 0.111 Jarque-Bera (JB): 4.479
Skew: -0.349 Prob(JB): 0.107
Kurtosis: 2.774 Cond. No. 4.61
==============================================================================
Warnings:
[1] Standard Errors are heteroscedasticity robust (HC3)
You can’t perform that action at this time.