generate_example_data: Generate example data

In marton-balazs-kovacs/tppr: Shiny app to show real-time results of the TPP project

Description

This code generates realistic example data for the Transparent Psi Project.

Usage

generate_example_data(
  num_sim = 6000,
  erotic_trial_size_per_participant = 18,
  m0_prob = 0.5,
  m1_prob = 0.51,
  esp_user_percentage = 1,
  chance_for_refuse_consent = 0.01,
  chance_for_stopping_session = 0.002
)

Arguments

num_sim	number of simulated sessions (1 regular session contains 36 trials, 18 erotic and 18 non-erotic)
erotic_trial_size_per_participant	number of erotic trials performed per participant
m0_prob	probability of successful guess if M0 is true
m1_prob	probability of successful guess among ESP-users in the simulated sample
esp_user_percentage	the percentage of ESP-users, or ESP-capable individuals in the population
chance_for_refuse_consent	chance of refusing consent or being excluded due to one of the eligibility criteria (e.g. underage)
chance_for_stopping_session	chance of stopping prematurely in each trial, to simulate some missing data due to unexpected events

Value

A dataframe containing the same data structure as expected from real data collected in the study.

More about the params

Set esp_user_percentage to 1 to simulate that everyone has the same level of ESP ability, or set it to 0 to simulate that no one has ESP ability. Set the value to something in between to simulate that only a portion of the population can use ESP, and the others are just guessing randomly. E.g., example esp_user_percentage = 0.03 and m1_prob = 0.65 simulates that only 3 of the total population can use ESP, and prediction success rate in this subgroup is 65 Setting chance_for_stopping_session to 0.002 would generate roughly 2 but that cannot introduce bias in our confirmatory analysis.

marton-balazs-kovacs/tppr documentation built on Oct. 27, 2021, 3:04 p.m.