Skip to content

wyattmcdonnell/sim_strength

Repository files navigation

A simple simulator for Gallup/StrengthsFinder 2.0 data

sim_strength is a simple simulator for understanding how a dataset of Clifton StrengthsFinder group frequencies (group_data.csv) is distributed relative to observed strength frequencies observed in a reference sample of ~12M Americans circa 2018.

Assumptions

  • We assume that choosing a top 5 is not very different from choosing a top 10
    • Several traits tend to cluster (e.g. "Relationship Building", "Influencing", etc.)
    • Some traits are frequently observed with each other (Intellection, Input; 50.9%)
    • Some traits are rarely observed with each other (Ideation, Discipline; 0.5%)
  • We assume that the conditional probabilities of traits being co-observed in a random sample of "top 5" results of 250,000 Americans generalize to 10

Input files

group_data.csv and reference_data.csv should both have frequencies in the range {0:1}. Please note that the simulator rounds to the second digit (e.g. 0.375 becomes 0.38). As there are 34 traits, both of these files should have entries for 34 traits. Note that you may wish to use your own reference_data.csv file from 2019 frequencies reported by Gallup for many countries. The reference_data.csv file in this repository refers to the American frequencies reported by Gallup.

probability_matrix.csv provides a conditional probability matrix of observing each pair of traits as reported in 250,000 random observations of American respondents in 2019.

Usage

Usage: cargo run -- <reference_data.csv> <group_data.csv> <group_size> <num_simulations> <verbose> <mode>

The program will run num_simulations of size group_size and report observed simulated frequencies for each of the 34 traits, and will report some simple statistical testing as well. Note that providing true for the verbose flag will print the results of each simulation to stdout. Consider the following sample command:

cargo run --release reference_data.csv group_data.csv 10 10000 false top5

The above command will run 10000 simulations of a group of 10 individuals, choose 5 top traits/strengths for each observation in each simulation, print a progress bar, and then report results.

Interpretation

Interpret at your own risk. Here be dragons! 🐉

About

simple Rust simulator for Gallup data

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages