Skip to content

cdsbasel/temprisk

BranchesTags

Repository files navigation

Meta-Analyses of the Temporal Stability and Convergent Validity of Risk Preference Measures

This repository contains the processed data (i.e., test-retest correlations and inter-correlations from the primary data sources) and code to replicate the analyses reported in the manuscript Meta-Analyses of the Temporal Stability and Convergent Validity of Risk Preference Measures.

Below we briefly describe the study (see Abstract), how this repository is organised (see Organization), how to replicate the analyses (see Scripts), and provide a description of the structure and content of the different data files created/used within the processing and analysis steps (see Codebook).

Additional information on the data sets, analyses, and results is available on the companion website.

Abstract

Understanding whether risk preference represents a stable, coherent trait is central to efforts aimed at explaining, predicting, and preventing risk-related behaviours. We help characterise the nature of the construct by adopting an individual participant data meta-analytic approach to summarise the temporal stability of over 350 risk preference measures (33 panels, 57 samples, >575,000 respondents). Our findings reveal significant heterogeneity across and within measure categories (propensity, frequency, behaviour), domains (e.g., investment, occupational, alcohol consumption), and sample characteristics (e.g., age). Specifically, while self-reported propensity and frequency measures of risk preference show a higher degree of stability relative to behavioural measures, these patterns are moderated by domain and age. Crucially, an analysis of convergent validity reveals a low agreement across measures, questioning the idea that they capture the same underlying phenomena. Our results raise concerns about the coherence and measurement of the risk preference construct.

Organization

  • var_info:

    • indv_panel_var_info:
      • PANEL_risk_var_info.csv: information on the variables from each panel/sample that are used for pre-processing.
      • PANEL_VariableInfo.xlsx: information on the risk preference measures from each panel/sample that are used for data analysis. Refer to the Codebook section for a description of the structure of the different files.
    • risk_measure_codebook.csv: description (incl. full wording of questions/items and options) of risk preference measures considered for analysis
    • panel_risk_info.rds: complete list of information on the risk preference measures for each panel/sample
    • panel_variable_info.rds: complete list of information on the variables for each panel/sample
    • code: scripts that combine all .xlsx or .csv files from the indv_panel_var_info folder to create the .rds files

  • pre_processing

    • code: scripts to select and pre-process the variables of interest from the raw data of each panel/sample.

  • processing

    • code:

      • temp_stability: scripts to compute test-retest correlations using the pre-processed data of each panel/sample.
      • convergent_val: scripts to compute inter-correlations using the pre-processed data of each panel/sample.

    • output:

      • temp_stability: .csv files with the test-retest correlations of all panels/samples combined as well as separate.
      • convergent_val: .csv files with the inter-correlations of of all panels/samples combined as well as separate.

      Refer to the Codebook section for a description of the columns in the different files.

  • analysis

    • code:

      • temp_stability: scripts to conduct the analyses on the temporal stability of risk preference measures.
      • convergent_val: scripts to conduct the analyses on the convergent validity of risk preference measures.

    • output:

      • temp_stability: output of the analyses on the temporal stability of risk preference measures.
      • convergent_val: output of the analyses on the convergent validity of risk preference measures.

      Refer to the Codebook section for a description of the columns in the different output (.csv) files.

  • plotting

    • code:
      • temp_stability: scripts to plot the test-retest correlations and analysis results.
      • convergent_val: scripts to plot the inter-correlations and analysis results.
    • output:
      • temp_stability: .png files of figures included in the manuscript.
      • convergent_val: .png files of figures included in the manuscript.

  • docs: files (.rmd and imagaes) for the companion website

  • main_data_files_codebook.xlsx: separate sheets containing the information on the structure and contents of main data files (Information also displayed in the Codebook section)

  • helper_functions.R: set of custom functions used for the processing and analysis of the data.

  • temprisk_info_session.txt: Information on the R environment, installed packages and their versions used for the analyses.

Scripts

A detailed description on how to run the scripts, on the output obtained, and on how to replicate the analyses reported in the manuscript can be found on the companion website. Additionally, each script contains a "Description" section. Lastly, refer to the temprisk_info_session.txt file for information on the packages required for the analyses.

Codebook

Description of the structure and content of the different data files created/used within the processing and analysis steps. Information also found in main_data_files_codebook.xlsx.

Analysis
Processing
Variable Info

VariableInfo_AND_risk_var_info

  • filename(s): PANEL_VariableInfo.xlsx and PANEL_risk_var_info.csv

  • file description: Files containing information on the variables used for pre-processing and the risk preference measures included in the analyses.

  • location: var_info/indv_panel_var_info/

column description type
panel Name of the panel character
wave_id ID of the wave character
varfile Name of the file containing the data (original name when the file was downloaded) character
origin_varcode Variable code in the original data character
varcode Standardized variable code character
measure_category Measure category of the variable (pro, fre, beh) character
general_domain Domain-general or domain-specific variable (gen or dom) character
domain_name Abbreviated mame of the domain of the variable (e.g., smo, alc, inv, gen) character
scale_type Type of response scale: ordinal (categorical variable with options that can be ranked), discrete (counts with clear range of possible responses, e.g., days in a month 0-30), open-ended (counts with no clear range), composite measure (sum of scores, proportions) character
scale_length If ordinal or discrete, the number of options/possible responses numeric
time_frame For frequency measures, the number of days the measure enquires about. numeric
behav_type For behavioural measures, the format of the task: lotteries, multiple price lists, willingness to pay/sell, allocation, dynamic character
behav_paid For behavioural measures, if it was incentivized or hypothetical character
check_var Has value 1 if it is a filter question in the survey numeric
item_num Number of items included in the measure numeric
comment comment character



risk_measure_codebook

  • filename(s): risk_measure_codebook.csv

  • file description: Main file containing the detailed description of the risk preference measures included in the analyses.

  • location: var_info/

column description type
panel Name of the panel character
measure_category Name of measure category (Propensity, Behavioural, Frequency) character
general_domain Whether it is general or domain-specific item character
domain_name Name of the domain (e.g., smoking, alcohol, investment) character
scale_type Type of scale (e.g., ordinal, open-ended) character
scale_length Number of response options (except for open-ended or composite measures) numeric
item_num Number of items included in the measure numeric
time_frame For frequency measures, the number of days the measure enquires about. numeric
survey_item How was the item phrased (copy-pasted from the panel questionnaire/codebook) character
response_options Reponse options (copy-pasted from the panel questionnaire/codebook) character
original_varcode Code of the variable in the panel codebook character
varcode Standardized varcode label character
info_source Where was information on the survey item collected from character
comment Additional comment (e.g., on which waves was this item included) character
reverse_coding In the pre-processing stage, do we need to reverse code the responses such that higher scores indicate more risk taking? (Y(es) or N(o)) character
behav_type For behavioural measures, the format of the task: lotteries, multiple price lists, willingness to pay/sell, allocation, dynamic character
behav_paid For behavioural measures, if it was incentivized or hypothetical character
dependencies Other variables to account for when pre-processing the data character



intercor_data

  • filename(s): PANEL/complete_intercor_data.csv

  • file description: Files containing the intercorrelations between risk preference measures

  • location: processing/output/convergent_val/

column description type
panel name of panel character
sample name of sample character
wave_id ID of wave character
wave_year Year in wich data collection took place numeric
year_age_group Size of age bins (i.e., 5, 10 or 20) numeric
age_group Age group of respondents (e.g., "10-19") character
age_mean Mean age of respondents numeric
age_median Median age of respondents numeric
age_min Minimum age of respondents numeric
age_max Maximum age of respondents numeric
age_sd Standard deviation of age of respondents numeric
gender_group Gender of respondents (i.e., female, male, all) character
prop_female Proportion of female respondents numeric
n Sample size numeric
varcode_a Name of variable A character
varcode_b Name of variable B character
cor_pearson Pearson correlation between variable A and B numeric
cor_spearman Spearman correlation between variable A and B numeric
icc2_1 ICC between variable A and B numeric
cor_pearson_log Pearson correlation between variable A and B for log-transformed responses numeric
cor_spearman_log Spearman correlation between variable A and B for log-transformed responses numeric
icc2_1_log ICC between variable A and B for log-transformed responses numeric
coeff_var_a Coefficient of variation for variable A responses numeric
coeff_var_b Coefficient of variation for variable B responses numeric
skewness_a Skewness of variable A responses numeric
skewness_b Skewness of variable B responses numeric
measure_category_a Measure category of variable A (pro, fre, beh) character
general_domain_a Whether variable A is a domain general or specific measure character
domain_name_a Domain of variable A (e.g., smo, alc) character
scale_type_a Type of response scale of variable A: ordinal (categorical variable with options that can be ranked), discrete (counts with clear range of possible responses, e.g., days in a month 0-30), open-ended (counts with no clear range), composite measure (sum of scores, proportions) character
scale_length_a If variable A is ordinal or discrete, the number of options/possible responses numeric
time_frame_a If variable A is a frequency measure, the number of days the measure enquires about. numeric
behav_type_a If variable A is a behavioural measure, the format of the task: lotteries, multiple price lists, willingness to pay/sell, allocation, dynamic character
behav_paid_a If variable A is a behavioural measure, if it was incentivized or hypothetical character
item_num_a Number of items included in variable A numeric
measure_category_b Measure category of variable B (pro, fre, beh) character
general_domain_b Whether variable B is a domain general or specific measure character
domain_name_b Domain of variable B (e.g., smo, alc) character
scale_type_b Type of response scale of variable B: ordinal (categorical variable with options that can be ranked), discrete (counts with clear range of possible responses, e.g., days in a month 0-30), open-ended (counts with no clear range), composite measure (sum of scores, proportions) character
scale_length_b If variable B is ordinal or discrete, the number of options/possible responses numeric
time_frame_b If variable B is a frequency measure, the number of days the measure enquires about. numeric
behav_type_b If variable B is a behavioural measure, the format of the task: lotteries, multiple price lists, willingness to pay/sell, allocation, dynamic character
behav_paid_b If variable B is a behavioural measure, if it was incentivized or hypothetical character
item_num_b Number of items included in variable B numeric
continent Continent where data collection took place character
country Country where data collection took place character
language Language of survey character
data_collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
sample_type Type of population who partakes in the survey: adolescents, adults, older adults, lifespan character



agg_intercorr

  • filename(s): agg_intercor_data.csv

  • file description: Files containing the aggregated intercorrelations between risk preference measures

  • location: processing/output/convergent_val/

column description type
panel Name of panel character
sample Name of sample character
continent Continent where data collection took place character
country Country where data collection took place character
language Language of survey character
data_collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
sample_type Type of population who partakes in the survey: adolescents, adults, older adults, lifespan character
age_group Age group of respondents (e.g., "10-19") character
gender_group Gender of respondents (i.e., female, male, all) character
meas_pair_lbl Measure category pair label (e.g., propensity-frequency) character
domain_pair_lbl Measure category-domain pair label (e.g., Propensity-General_Frequency-Smoking) character
n_mean Mean sample size of correlations numeric
n_sd Standard deviation of sample sizes of the correlations numeric
mean_age Mean age of the respondents (i.e., mean of the mean age of respondents ) numeric
sd_age Standard deviation of the mean age of the respondents numeric
cor_num Number of correlations included to calculate the aggregate estimate numeric
wcor_z Fisher's z values of the aggregated inter-correlation numeric
vi_z Sampling variance of Fisher's z aggregated estimate numeric
sei_z Square root of vi_z numeric
ci_lb_z Lower 95% bound of Fisher's z aggregate numeric
ci_ub_z Upper 95% bound of Fisher's z aggregate numeric
wcor z-to-r transformed aggregated inter-correlation numeric
ci_lb Lower 95% bound of r inter-correlation aggregate numeric
ci_ub Upper 95% bound of r inter-correlation aggregate numeric
sei Square root of vi numeric
vi Sampling variance of r aggregated estimate numeric
es_id ID of effect size numeric
age_bin Age binning (5, 10, or 20-year bins) numeric
min_n Minimum sample size of correlations included to compute the aggregated estimates numeric
rho_val Correlation between sampling errors of effect sizes being aggregated numeric
data_transform Whether correlations were computed from the non or log-transformed responses character
cor_metric Correlation metric: pearson, spearman, ICC character



retest_data

  • filename(s): PANEL/complete_retest_data.csv

  • file description: Files containing the retest correlation for each risk preference measure

  • location: processing/output/temp_stability/

column description type
panel name of panel character
sample name of sample character
wave_id_t1 ID of wave at T1 character
wave_year_t1 Year in wich data collection took place at T1 numeric
wave_id_t2 ID of wave at T2 character
wave_year_t2 Year in wich data collection took place at T2 numeric
time_diff_mean Mean time difference (in years) between T1 and T2 numeric
time_diff_median Median time difference (in years) between T1 and T2 numeric
time_diff_min Minimum time difference (in years) between T1 and T2 numeric
time_diff_max Maximum time difference (in years) between T1 and T2 numeric
time_diff_sd Standard deviaton of time difference (in years) between T1 and T2 numeric
year_age_group Size of age bins (i.e., 5, 10 or 20) numeric
age_group Age group of respondents (e.g., "10-19") character
age_mean Mean age of respondents numeric
age_median Median age of respondents numeric
age_min Minimum age of respondets numeric
age_max Maximum age of respondenrts numeric
age_sd Standard deviation of age of respondents numeric
gender_group Gender of respondents (i.e., female, male, all) character
prop_female Proportion of female respondents numeric
n Number of respondents numeric
attrition_rate Proportion of T1 respondents missing at T2 numeric
varcode Variable code of measure character
cor_pearson Pearson correlation between responses at T1 and T2 numeric
cor_spearman Spearman correlation between responses at T1 and T2 numeric
icc2_1 ICC between responses at T1 and T2 numeric
cor_pearson_log Pearson correlation between responses at T1 and T2 for log-transformed responses numeric
cor_spearman_log Spearman correlation between responses at T1 and T2 for log-transformed responses numeric
icc2_1_log ICC between responses at T1 and T2 for log-transformed responses numeric
coeff_var_t1 Coefficient of variation for variable T1 responses numeric
coeff_var_t2 Coefficient of variation for variable T2 responses numeric
skewness_t1 Skewness of variable T1 responses numeric
skewness_t2 Skewness of variable T2 responses numeric
measure_category Measure category of the variable (pro, fre, beh) character
general_domain Domain-general or domain-specific variable (gen or dom) character
domain_name Name of domain of variable (e.g., smo, alc) character
scale_type Type of response scale: ordinal (categorical variable with options that can be ranked), discrete (counts with clear range of possible responses, e.g., days in a month 0-30), open-ended (counts with no clear range), composite measure (sum of scores, proportions) character
scale_length If ordinal or discrete, the number of options/possible responses numeric
time_frame For frequency measures, the number of days the measure enquires about. numeric
behav_type For behavioural measures, the format of the task: lotteries, multiple price lists, willingness to pay/sell, allocation, dynamic character
behav_paid For behavioural measures, if it was incentivized or hypothetical character
item_num Number of items included in the measure numeric
continent Continent where data collection took place character
country Country where data collection took place character
language Language of survey character
data_collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
sample_type Type of population who partakes in the survey: adolescents, adults, older adults, lifespan character



agg_retest

  • filename(s): agg_retest_data.csv

  • file description: Files containing aggregated retest correlations

  • location: processing/output/temp_stability/

column description type
panel Name of panel character
sample Name of sample character
continent Continent where data collection took place character
country Country where data collection took place character
language Language of survey character
data_collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
sample_type Type of population who partakes in the survey: adolescents, adults, older adults, lifespan character
time_diff_bin Rounded mean time difference between T1 and T2 numeric
age_group Age group of respondents (e.g., "10-19") character
gender_group Gender of respondents (i.e., female, male, all) character
measure_category Measure category of the variable (pro, fre, beh) character
domain_name Name of domain of variable (e.g., smo, alc) character
item_num Whether measures are "single item" or "multi item" measures character
n_mean Mean sample size of correlations numeric
n_sd Standard deviation of sample sizes of the correlations numeric
mean_age Mean age of the respondents (i.e., mean of the mean age of respondents ) numeric
sd_age Standard deviation of the mean age of the respondents numeric
mean_attrition Mean attrition rate between T1 and T2 numeric
sd_attrition Standard deviation of attrition rate between T1 and T2 numeric
cor_num Number of correlations included to calculate the aggregate estimate numeric
wcor_z Fisher's z value of the aggregated retest correlation numeric
vi_z Sampling variance of Fisher's z aggregated estimate numeric
sei_z Square root of vi_z numeric
ci_lb_z Lower 95% bound of Fisher's z aggregate numeric
ci_ub_z Upper 95% bound of Fisher's z aggregate numeric
wcor z-to-r transformed aggregated retest correlation numeric
ci_lb Lower 95% bound of r retest aggregate numeric
ci_ub Upper 95% bound of r retest aggregate numeric
sei Square root of vi numeric
vi Sampling variance of aggregated estimate numeric
es_id Effect size id numeric
age_bin Age binning (5, 10, or 20-year bins) numeric
min_n Minimum sample size of correlations included to compute the aggregated estimates (30, 100, or 250) numeric
month_bin Binning of time difference (3, 6, or 12-month bins) numeric
rho_val Correlation between sampling errors of effect sizes being aggregated numeric
data_transform Whether correlations were computed from the non or log-transformed responses character
cor_metric Correlation metric: pearson, spearman, ICC character



complete_retest_info

  • filename(s): complete_retest_info.csv

  • file description: File containing summary information on the amount of data and number of retest correlations analysed in each sample (used to create the flowchart)

  • location: processing/output/temp_stability/

column description type
panel Name of panel character
sample Name of sample character
continent Continent where data collection took place character
country Country where data collection took place character
sample_type Type of population who partakes in the survey: adolescents, adults, older adults, lifespan character
collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
measure_categ List of the categories of the measures anaylsed (pro, fre, beh) character
domain_name List of the domains of the measures anaylsed (e.g., smo, alc) character
unique_meas Number of unique measures numeric
unique_waves Number of waves numeric
retest_int_min Minimum retest interval (years) numeric
retest_int_median Median retest interval (years) numeric
retest_int_mean Mean retest interval (years) numeric
retest_int_max Maximum retest interval (years) numeric
cor_num Number of correlations analysed numeric
unique_id Number of unique respondents numeric
unique_resp Number of unique responses numeric



complete_intercor_info

  • filename(s): complete_intercor_info.csv

  • file description: File containing summary information on the amount of data and number of intercorrelations analysed in each sample (used to create the flowchart)

  • location: processing/output/convergent_val/

column description type
panel Name of panel character
sample Name of sample character
continent Continent where data collection took place character
country Country where data collection took place character
collect_mode Mode of data collection used across most waves: interview, online, laboratory, self-administered (e.g., survey sent via post) character
measure_categ List of the categories of the measures analysed (pro, fre, beh) character
domain_name List of the domains of the measures analysed (e.g., smo, alc) character
unique_meas Number of unique measures numeric
cor_num Number of correlations analysed numeric
unique_id Number of unique respondents numeric
unique_resp Number of unique responses numeric



temp_stability-masc_nlpar_pred

  • filename(s): temp_stability/masc_nlpar_pred.csv

  • file description: Summary values of MASC parameter estimates for risk preference and other psych constructs (plot to compare constructs, predictor of interest is domain, all other predictors are set to 0 )

  • location: analysis/output/temp_stability/

column description type
categ Name of the predictor (e.g., domain,) character
x Name of the predictor level (e.g., smo, inv) character
measure Measure category (Propensity, Behaviour, Frequency) character
nlpar Name of the non linear parameter character
.epred Mean value of the paramter estimate numeric
.lower_0.95 Value of the lower 95% HDI of the parameter estimate numeric
.lower_0.8 Value of the lower 80% HDI of the parameter estimate numeric
.lower_0.5 Value of the lower 50% HDI of the parameter estimate numeric
.upper_0.95 Value of the upper 95% HDI of the parameter estimate numeric
.upper_0.8 Value of the upper 80% HDI of the parameter estimate numeric
.upper_0.5 Value of the upper 50% HDI of the parameter estimate numeric
sub_component Relabeled x variable character



convergent_val-masc_nlpar_pred

  • filename(s): convergent_val/masc_nlpar_pred.csv

  • file description: Summary ofMASC parameter estimates for risk preference for different predictor values (used for variance decomp. analysis)

  • location: analysis/output/convergent_val/

column description type
categ Name of the predictor (e.g., domain,) character
x Name of the predictor level (e.g., smo, inv) character
measure Measure category (Propensity, Behaviour, Frequency) character
age_group Age group (e.g., "20-30") character
gender_group Gender group (male, female) character
nlpar Name of the non-linear parameter character
.epred Mean value of the parameter estimate numeric
.lower_0.95 Value of the lower 95% HDI of the parameter estimate numeric
.lower_0.8 Value of the lower 80% HDI of the parameter estimate numeric
.lower_0.5 Value of the lower 50% HDI of the parameter estimate numeric
.upper_0.95 Value of the upper 95% HDI of the parameter estimate numeric
.upper_0.8 Value of the upper 80% HDI of the parameter estimate numeric
.upper_0.5 Value of the upper 50% HDI of the parameter estimate numeric
sub_component Relabeled x variable character



shapley_values_boot

  • filename(s): shapley_values_measure_retest_boot.csv AND shapley_values_intercor_boot.csv

  • file description: Variance Decomposition output for bootstrapped samples including the R2 values of including and excluding specific predictors

  • location: analysis/output/temp_stability/ AND analysis/output/covergent_val/

column description type
r2_increment Difference between of r2_with and r2_without numeric
r2_with Value of R2 by including the predictor of interest numeric
r2_without Value of R2 by excluding the predictor of interest numeric
r2adj_increment Difference between of r2adj_with and r2adj_without numeric
r2adj_with Value of adjusted R2 by including the predictor of interest numeric
r2adj_without Value of adjusted R2 by excluding the predictor of interest numeric
x Predictor of interest (e.g., age, domain) character
boot_num Bootstrapped sample number numeric
n_reg_with Number of predictors in the model numeric
supple List of predictors in the model character



shapley_values

  • filename(s): shapley_values_measure_retest.csv and shapley_values_intercor.csv

  • file description: Variance Decomposition output for the dataset including the R2 values of including and excluding specific predictors

  • location: analysis/output/temp_stability/ and analysis/output/convergent_val/

column description type
row_id Row number numeric
r2_increment Difference between of r2_with and r2_without numeric
r2_with Value of R2 by including the predictor of interest numeric
r2_without Value of R2 by excluding the predictor of interest numeric
r2adj_increment Difference between of r2adj_with and r2adj_without numeric
r2adj_with Value of adjusted R2 by including the predictor of interest numeric
r2adj_without Value of adjusted R2 by excluding the predictor of interest numeric
x Predictor of interest (e.g., age, domain) character
n_reg_with Number of predictors in the model numeric
supple List of predictors in the model character



shapley_values_check

  • filename(s): shapley_values_check.csv

  • file description: Checking for issues of singularity in the variance decomposition analysis

  • location: analysis/output/temp_stability/ and analysis/output/convergent_val/

column description type
check Issue of singularity in regression? logical



summary_shapley_values

  • filename(s): summary_shapley_values_retest.csv AND summary_shapley_values_intercor.csv

  • file description: Summarised Variance Decomposition output for plotting

  • location: analysis/output/temp_stability/ AND analysis/output/convergent_val/

column description type
x Predictor of interest character
measure_category Name of measure category (Behaviour, Frequency, Propensity, Omnibus) character
m Shapley Value (i.e., weighted adjusted R2 increment) numeric
x_lbl Relabeled predictor name for plotting character
categ_lbl Category/Family of predictors the predictor belongs to (i.e., panel, respondent, measure) character



summary_shapley_values_boot

  • filename(s): summary_shapley_values_retest_boot.csv AND summary_shapley_values_intercor_boot.csv

  • file description: Summarised Variance Decomposition (boostrapped) output for plotting

  • location: analysis/output/temp_stability/ AND analysis/output/convergent_val/

column description type
x Predictor of interest character
measure_category Name of measure category (Behaviour, Frequency, Propensity, Omnibus) character
m Overall mean of Shapley Values (i.e., weighted adjusted R2 increment) accoss all boostrapped samples numeric
.point Mean character
.interval Quantiles character
.lower_0.5 Lower 50th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
.lower_0.8 Lower 80th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
.lower_0.95 Lower 95th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
.upper_0.5 Upper 50th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
.upper_0.8 Upper 80th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
.upper_0.95 Upper 95th quantile of Shapley Values (i.e., weighted adjusted R2 increment) across all boostrapped samples numeric
x_lbl Relabeled predictor name for plotting character
categ_lbl Category/Family of predictors the predictor belongs to (i.e., panel, respondent, measure) character



cor_mat_convergent

  • filename(s): cor_mat_convergent_.csv

  • file description: Convergent Validity data for plotting

  • location: analysis/output/convergent_val/

column description type
param Name of model regression from the regression character
estimate Meta-Analytic estimate for the intercorrelation numeric
.lower Lower 95% HDI of estimate numeric
.upper Upper 95% HDI of estimate numeric
.width 95 numeric
.point mean character
.interval hdci character
meas_pair_id ID of measure pairs numeric
meas_pair_lbl Label of measure pairs character
x x-axis label/text for plotting character
y y-axis label/text for plotting character
n_cor Number of "raw" correlations numeric
n_wcor Number of aggregated correlations that were analysed numeric
pooled_est_lbl Label of pooled estimate for plotting character
cred_int_lbl Label of upper and lower HDCI for plotting character
k_lbl Label of n_wcor for plotting character
lbl_color Color of the text label for the plotting character



About

No description, website, or topics provided.

Resources

Readme

License

GPL-3.0 license
Activity
Custom properties

Stars

0 stars

Watchers

1 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages