model_outputs.md

jupytext	kernelspec
formats text_representation ipynb,md:myst extension format_name format_version jupytext_version .md myst 0.13 1.14.5	display_name language name Python 3 (ipykernel) python python3

import json

from survey_analysis import format_data, fit_model

responses = format_data('public_survey_data.csv')

# now let's run some basic analyses to check against SurveyMonkey outputs
responses['is_member'].value_counts(normalize=True)

responses['geographic_region'].value_counts(normalize=True)

responses['career_stage'].value_counts(normalize=True)

# for the other questions, we can load in our sidecar JSON
# with question-level metadata
with open('levels.json') as f:
    queries = json.load(f)

print(f'There are {len(queries.keys())} questions to consider.') 
print(queries.keys())

question = 'job_board_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'job_board_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'email_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'email_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'blog_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'blog_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'twitter_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'twitter_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'podcast_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'podcast_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'facebook_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'youtube_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'youtube_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'linkedin_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'linkedin_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'ondemand_access'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'ondemand_content'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'content_platform'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'sig_platform'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())

question = 'content_importance'
scale_values = queries[question]

res = fit_model(responses, question, scale_values)
print(res.summary())