The Challenger Space Shuttle Disaster: A Case Study in the Analysis of Binary Data Using Scatter Diagrams and Logit Regression

• Published date: 01 June 2021
• Author: Robert Dixon
• Date: 01 June 2021
• DOI: http://doi.org/10.1111/1467-8462.12410

The Australian Economic Review, vol. 54, no. 2, pp. 294–305 DOI: 10.1111/1467-8462.12410

For the student

The Challenger Space Shuttle Disaster: A Case Study in the Analysis

of Binary Data Using Scatter Diagrams and Logit Regression

Robert Dixon*

Abstract

This article offers an instructive example of

the usefulness of scatter diagrams and of logit

regression in decision making. On the

morning of 28 January 1986 a US space

shuttle named ‘Challenger’exploded soon

after take‐off killing all seven crew members

who were on board. The launch went ahead

on a very cold morning despite attempts by a

group of engineers to halt it. They argued that

it was unsafe to proceed given the likelihood

of malfunctions occurring when the tempera-

ture was below freezing. They failed to

convince the decision makers involved of

any strong connection between temperature

at launch and the malfunction of a particular

crucial part. This articles discusses how the

engineers’proper use of scatter diagrams

and/or logit regression would likely have

supported a convincing case to delay the

launch of the shuttle.

1. Introduction

The aim of this article is to show a very

instructive example of the usefulness of

scatter diagrams and of logit regression in

decision making. It uses the Challenger Space

Shuttle disaster as a case study.

I begin by summarising the circumstances in

which on the morning of 28 January 1986 a US

space shuttle named ‘Challenger’exploded soon

after take‐off killing all seven crew members

who were on board. The cause of the accident

was found to be the failure of a rubber seal

(called an ‘O‐ring’) which was in one of the

shuttle's rocket boosters. Evidence was given to

an inquiry into the disaster that the O‐rings were

liable to malfunction in cold weather and that at

the time of launch the temperature was below

freezing. On the night before the Challenger was

launched a large group of engineers and

managers connected with the Shuttle program

met for a number of hours to discuss some

concerns about the consequences of the forecast

record low temperatures and the ability of the

O‐rings to function properly. Although

the engineers expressed their opposition to the

launch, at the conclusion of the meeting the

managers decided that the engineers had not

made a sufficiently convincing case that seal

failure was related to temperature and, conse-

quently, decided to continue with the launch of

the shuttle with catastrophic results. It is widely

thought that this mistake would not have

occurred had the engineers and managers

collected and studied all relevant (and readily

* Department of Economics, the University of

Melbourne. Corresponding author: Dixon, email: <r.dix-

on@unimelb.edu.au>. I am grateful to Jenny Lye for

helpful comments on an earlier version of the article. I am

also grateful to the Office of National Archives in

Washington DC, for permission to reprint material from

the 1986 Report of the Presidential Commission on the

Space Shuttle Challenger Accident, Volume 1.

© 2021 The University of Melbourne, Melbourne Institute: Applied Economic & Social Research,

Faculty of Business and Economics

Published by John Wiley & Sons Australia, Ltd