Harwell, M. and Gatti. G. (2001) ‘Rescaling ordinal data into interval data in educational research’, Review of Educational Research, Spring, 71 (1): 105–31.

These authors argue that in educational research, many statistical procedures used require that dependent variables are metric (‘interval scale of measurement’ in their terminology) but in practice are only ordered category. The authors go on to suggest a technique called item response theory for rescaling ordinal variables into metric ones, so this article is relevant for both this chapter and for Chapter 2, which considers data transformation processes

Hox, J. de Leeuw, E. and Chang, H., (2012) ‘Nonresponse versus measurement error: are reluctant respondents worth pursuing?’, Bulletin de Methodologie Sociologique, 113: 5–19.

This article on error in data construction is well worth reading. The early part of the article in particular is comparing non-response and measurement error.

Sandelowski, M , Voils, C. and Knafl, G. (2009) ‘On quantitizing’, Journal of Mixed Methods Research, 3 (3): 208–22.

‘Quantitizing’ means translating, transforming or converting qualitative data into quantitative. Although the article is written in the context of mixed methods, there is a discussion of the nature of data and the assumptions made in creating numerical data. In particular, judgements have to be made about what and how to count if frequencies are to be created.