Answers are grounded in this paper and cite their sources.
With large-scale data collection stimulated by computer analysis the problem of what to do about missing observarions on Ss arises more often. In a research project by Turner, Guertin, and Cooper' correlations of GPA with 10 achievement rest scores were computed three ways because of missing scores for some Ss: ( a ) insert the mean score for that variable, ( b ) insert a multiple regression estimate, and ( c ) omit that S from the calculation. Results are given in the table, with the per cent of missing observations given in italics.
No takes yet. Share an insight, caveat, or question.
Wilson H. Guertin (1968) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: