Abstract
| Developed at the turn of the 20th century, factor analysis is often cited as the
most proliferate of psychometric techniques. Yet proof was never secured for the
mathematical foundations of factor analysis by original inventor Charles
Spearman. Factor analysis later was transformed into a statistical method
drawing on Ronald Fisher's maximum likelihood estimation, yet similarly,
although sought, mathematical proof for this technique was never found. In this
talk we examine what this lack of mathematical proof means for factor analysis
as a realist theory of measurement from a logical perspective, and we explore
the work of mathematician Penelope Maddy to discern how we might rescue factor
analysis by utilising a systematised axiomatic approach informed by the
development of set theory. |