The Galton Height Predictor

Submission for Coursera Data Science Project

rutulian

The Galton Height Predictor is a handy and convenient way to inaccurately predict the heights of any Victorian children to be born!

The dataset

The data used to create this predictor was collected by Francis Galton in 1885. It is a list of midparent heights and child heights, both in inches. The midparent height is calculated as the average of the father's height and 1.08 * mother's height.

The first 6 items:

child parent
1 61.70 70.50
2 61.70 68.50
3 61.70 65.50
4 61.70 64.50
5 61.70 64.00
6 62.20 67.50

A summary of the data:

child parent
1 Min. :61.70 Min. :64.00
2 1st Qu.:66.20 1st Qu.:67.50
3 Median :68.20 Median :68.50
4 Mean :68.09 Mean :68.31
5 3rd Qu.:70.20 3rd Qu.:69.50
6 Max. :73.70 Max. :73.00

The predictive model

  1. The child height \(Y\) is the dependent variable
  2. The parent height \(X\) is the independent variable
  3. A simple linear regression of the form \(Y = \alpha + \beta X\) is performed
lm(galton)$coefficients

## (Intercept)      parent 
##  23.9415302   0.6462906

How to use

  1. Go to https://rutulian.shinyapps.io/GaltonShiny
  2. Enter the height of the father in feet and inches
  3. Enter the height of the mother in feet and inches
  4. Observe the histogram, which displays where the child of these two parents is predicted to lie within the distribution of heights of children in the original dataset

Please note, the heights of father and mother are restricted to be close to the extreme values within the original dataset.