uncongeniality_analysis/analysis_jobs/analysis_job_manuscript.yaml

---
preprocessing:
  data_order0:
  - method: data_order
    param: 0
  data_order1:
  - method: data_order
    param: 1
  data_politics:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Politics'
  data_foreign_affairs:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Foreign affairs'
  data_science:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Science'
  data_economy:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Economy'
  data_miscellaneous:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Miscellaneous'
  data_culture:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Culture'
  data_sports:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Sports'
  data_mobility:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Mobility'
  data_internet:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Internet'
  data_health:
  - method: data_order
    param: 0
  - method: data_section
    param: 'Health'
  data_order0_with_minimum_one_vote:
  - method: data_order
    param: 0
  - method: exclude_data_with_value
    param: {'column': 'totalvotes', 'value': 0}

descriptive:
  - !descriptive_overview
    name: "Extended_Data_Table_1_Descriptive_Data_for_different_comment_levels"
    dataset: "data"
    group_by: "order"
    metrics:
      - operation: "count"
        column: null
      - operation: "count_nonzero"
        column: "totalvotes"
      - operation: "sum"
        column: "totalvotes"
      - operation: "mean"
        column: "totalvotes"
      - operation: "std_dev"
        column: "totalvotes"
      - operation: "sum"
        column: "upvotes"
      - operation: "sum"
        column: "downvotes"
      - operation: "mean"
        column: "bayes-corrected (q=0.25) valence"
      - operation: "std_dev"
        column: "bayes-corrected (q=0.25) valence"
      - operation: "mean"
        column: "bayes-corrected (q=0.25) extremity"
      - operation: "std_dev"
        column: "bayes-corrected (q=0.25) extremity"

  - !descriptive_overview
    name: "Extended_Data_Table_2_Descriptive_Data_for_different_news_categories"
    dataset: "data"
    group_by: "section"
    metrics:
      - operation: "count"
        column: null
      - operation: "sum"
        column: "number O(n+1)-replies"
      - operation: "count_nonzero"
        column: "number O(n+1)-replies"
      - operation: "count_nonzero"
        column: "totalvotes"
      - operation: "sum"
        column: "totalvotes"
      - operation: "sum"
        column: "upvotes"
      - operation: "sum"
        column: "downvotes"
      - operation: "count_nonzero"
        column: "totalvotes"
      - operation: "mean"
        column: "valence"
      - operation: "std_dev"
        column: "valence"
      - operation: "mean"
        column: "bayes-corrected (q=0.25) valence"
      - operation: "std_dev"
        column: "bayes-corrected (q=0.25) valence"
      - operation: "mean"
        column: "extremity"
      - operation: "std_dev"
        column: "extremity"
      - operation: "mean"
        column: "bayes-corrected (q=0.25) extremity"
      - operation: "std_dev"
        column: "bayes-corrected (q=0.25) extremity"

analysis:
  - !linear_regression
    name: "Evidence_uncongeniality_simplest_model_linear_regression_only_valence_non_standardized"
    dataset: "data_order0"
    independent_variables:
      - 'valence'
    dependent_variable: 'number O(n+1)-replies'
    standardize: false
    report_effect_size: true

  - !linear_regression
    name: "Evidence_uncongeniality_preregistered_model"
    dataset: "data_order0"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true
    report_effect_size: true

  - !linear_regression
    name: "Evidence_uncongeniality_stability_against_variation_in_weight_q5"
    dataset: "data_order0"
    independent_variables:
      - 'bayes-corrected (q=0.5) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongeniality_stability_against_variation_in_weight_q75"
    dataset: "data_order0"
    independent_variables:
      - 'bayes-corrected (q=0.75) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongeniality_stability_against_variation_in_weight__no_bayes_correction"
    dataset: "data_order0"
    independent_variables:
      - 'valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression_grouped
    name: "Evidence_uncongeniality_robustness_analysis_on_person_level"
    dataset: "data_order0"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    aggregation_functions:
      - 'mean'
      - 'sum'
      - 'sum'
    group_by: 'user_id'
    standardize: true
    print_detailed_coefficients: true

  - !linear_regression_grouped
    name: "Evidence_uncongeniality_robustness_analysis_on_section_level"
    dataset: "data_order0"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    aggregation_functions:
      - 'mean'
      - 'sum'
      - 'sum'
    group_by: 'section'
    standardize: true
    print_detailed_coefficients: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_politics"
    dataset: "data_politics"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_affairs"
    dataset: "data_foreign_affairs"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_science"
    dataset: "data_science"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_economy"
    dataset: "data_economy"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_miscellaneous"
    dataset: "data_miscellaneous"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_culture"
    dataset: "data_culture"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_sports"
    dataset: "data_sports"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_mobility"
    dataset: "data_mobility"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_internet"
    dataset: "data_internet"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongenialty_section_health"
    dataset: "data_health"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncongeniality_robustness_order1"
    dataset: "data_order1"
    independent_variables:
      - 'bayes-corrected (q=0.25) valence'
      - 'totalvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_uncogeniality_model_with_seperate_upvotes_downvotes"
    dataset: "data_order0"
    independent_variables:
      - 'upvotes'
      - 'downvotes'
    dependent_variable: 'number O(n+1)-replies'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_preregistered_model"
    dataset: "data_order0"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_stability_against_variation_in_weight_q5"
    dataset: "data_order0"
    independent_variables:
      - 'mean bayes-corrected (q=0.5) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.5) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_stability_against_variation_in_weight_q75"
    dataset: "data_order0"
    independent_variables:
      - 'mean bayes-corrected (q=0.75) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.75) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_stability_against_variation_in_weight_no_bayes_correction"
    dataset: "data_order0"
    independent_variables:
      - 'mean valence of replies'
    dependent_variable: 'valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_politics"
    dataset: "data_politics"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_affairs"
    dataset: "data_foreign_affairs"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_science"
    dataset: "data_science"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_economy"
    dataset: "data_economy"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_miscellaneous"
    dataset: "data_miscellaneous"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_culture"
    dataset: "data_culture"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_sports"
    dataset: "data_sports"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_mobility"
    dataset: "data_mobility"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_internet"
    dataset: "data_internet"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_section_health"
    dataset: "data_health"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !linear_regression
    name: "Evidence_antagonism_robustness_order1"
    dataset: "data_order1"
    independent_variables:
      - 'mean bayes-corrected (q=0.25) valence of replies'
    dependent_variable: 'bayes-corrected (q=0.25) valence'
    standardize: true

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity"
    dataset: "data_order0"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_stability_against_variation_in_weight_paired_ttest_q5"
    dataset: "data_order0"
    variable_1: 'bayes-corrected (q=0.5) extremity'
    variable_2: 'mean bayes-corrected (q=0.5) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_stability_against_variation_in_weight_paired_ttest_q75"
    dataset: "data_order0"
    variable_1: 'bayes-corrected (q=0.75) extremity'
    variable_2: 'mean bayes-corrected (q=0.75) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_stability_against_variation_in_weight_paired_ttest_bayes"
    dataset: "data_order0"
    variable_1: 'extremity'
    variable_2: 'mean extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_robustness_paired_ttest_order1"
    dataset: "data_order1"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_politics"
    dataset: "data_politics"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_foreign_affairs"
    dataset: "data_foreign_affairs"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_science"
    dataset: "data_science"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_economy"
    dataset: "data_economy"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_miscellaneous"
    dataset: "data_miscellaneous"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_culture"
    dataset: "data_culture"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_sports"
    dataset: "data_sports"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_mobility"
    dataset: "data_mobility"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_internet"
    dataset: "data_internet"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

  - !paired_ttest
    name: "Evidence_polarization_paired_ttest_extremity_health"
    dataset: "data_health"
    variable_1: 'bayes-corrected (q=0.25) extremity'
    variable_2: 'mean bayes-corrected (q=0.25) extremity of replies'

visualization:
  - !hexbinplot
    name: "Fig_2a"
    dataset: "data_order0"
    variable_x_axis: 'bayes-corrected (q=0.25) valence'
    variable_y_axis: 'number O(n+1)-replies'
    y_axis_maximum: 40
    trendline: True
    logarithmic_hex_scaling: True

  - !forestplot
    name: "Fig_2b"
    regression_model_names:
      - "Evidence_uncongenialty_section_politics"
      - "Evidence_uncongenialty_section_foreign_affairs"
      - "Evidence_uncongenialty_section_science"
      - "Evidence_uncongenialty_section_economy"
      - "Evidence_uncongenialty_section_miscellaneous"
      - "Evidence_uncongenialty_section_culture"
      - "Evidence_uncongenialty_section_sports"
      - "Evidence_uncongenialty_section_mobility"
      - "Evidence_uncongenialty_section_internet"
      - "Evidence_uncongenialty_section_health"
    regression_model_labels:
      - "Politics"
      - "Foreign Affairs"
      - "Science"
      - "Economy"
      - "Miscellaneous"
      - "Culture"
      - "Sports"
      - "Mobility"
      - "Internet"
      - "Health"
    coefficient_names:
        - "bayes-corrected (q=0.25) valence"
        - "totalvotes"
    x_axis_minimum: -0.6
    dotsize: 2
    x_axis_label: "Standardized coefficient (95% Confidence Interval)"

  - !heatmap
    name: "Fig_2c"
    dataset: "data_order0_with_minimum_one_vote"
    axis_variables:
      - 'upvotes'
      - 'downvotes'
    heat_variable: 'number O(n+1)-replies'
    axis_maxima:
      - 20
      - 20
    axis_minima:
      - 0
      - 0
    logarithmic_heat_scaling: 'false'

  - !densityplot
    name: 'Fig_3a'
    dataset: "data_order0"
    variable_x_axis: 'mean bayes-corrected (q=0.25) valence of replies'
    variable_y_axis: 'bayes-corrected (q=0.25) valence'
    data_breakpoints:
      - 0

  - !forestplot
    name: "Fig_3b"
    regression_model_names:
      - "Evidence_antagonism_section_politics"
      - "Evidence_antagonism_section_foreign_affairs"
      - "Evidence_antagonism_section_science"
      - "Evidence_antagonism_section_economy"
      - "Evidence_antagonism_section_miscellaneous"
      - "Evidence_antagonism_section_culture"
      - "Evidence_antagonism_section_sports"
      - "Evidence_antagonism_section_mobility"
      - "Evidence_antagonism_section_internet"
      - "Evidence_antagonism_section_health"
    regression_model_labels:
      - "Politics"
      - "Foreign Affairs"
      - "Science"
      - "Economy"
      - "Miscellaneous"
      - "Culture"
      - "Sports"
      - "Mobility"
      - "Internet"
      - "Health"
    coefficient_names:
        - 'mean bayes-corrected (q=0.25) valence of replies'
    x_axis_minimum: -0.1
    dotsize: 2
    x_axis_label: "Standardized coefficient (95% Confidence Interval)"

  - !violinplot
    name: "Fig_4a"
    dataset: "data_order0"
    variable_x_axis: 'bayes-corrected (q=0.25) extremity'
    variable_y_axis: 'mean bayes-corrected (q=0.25) extremity of replies'
    x_axis_label: ''
    y_axis_label: 'Extremity value'
    title: ''

  - !forestplot_paired_ttest
    name: "Fig_4b"
    paired_ttest_names:
        - "Evidence_polarization_paired_ttest_extremity_politics"
        - "Evidence_polarization_paired_ttest_extremity_affairs"
        - "Evidence_polarization_paired_ttest_extremity_science"
        - "Evidence_polarization_paired_ttest_extremity_economy"
        - "Evidence_polarization_paired_ttest_extremity_miscellaneous"
        - "Evidence_polarization_paired_ttest_extremity_culture"
        - "Evidence_polarization_paired_ttest_extremity_sports"
        - "Evidence_polarization_paired_ttest_extremity_mobility"
        - "Evidence_polarization_paired_ttest_extremity_internet"
        - "Evidence_polarization_paired_ttest_extremity_health"
    paired_ttest_labels:
      - "Politics"
      - "Foreign Affairs"
      - "Science"
      - "Economy"
      - "Miscellaneous"
      - "Culture"
      - "Sports"
      - "Mobility"
      - "Internet"
      - "Health"
    x_axis_minimum: -0.06
    dotsize: 2
    x_axis_label: "Mean difference bayes-corrected (q=0.25) extremity (95% Confidence Interval)"

  - !histogram
    name: 'Extended_Fig_1'
    dataset: "data"
    variable: 'totalvotes'
    x_axis_label: 'Number of total votes'
    y_axis_label: 'Number of comments'
    x_axis_logarithmic_scaling: false
    y_axis_logarithmic_scaling: true
    title: ''
...