correlation analysis was also used to analyze the study's data. Canonical correlation is a
variation of multiple regression and correlation. In multiple regression analysis, the
relationship between a single variable and a linear combination of a set of variables is
examined. In the canonical correlation procedure, the relationship between a linear
combination of a set of variables and a linear combination of another set of variables is
analyzed. Rather than predicting a single dependent variable from a set of independent
variables, this procedure correlates two sets of variables.
Description of the Data
The survey used in this study consisted of four previously-established measures,
one assessment designed for this investigation, and a demographic questionnaire. The
results of the demographic questionnaire were presented in Chapter 3, where the sample
was described. The ranges, means, and standard deviations for each of the study variables
are presented in Table 4-1. In a few cases, participants gave more than one response to an
item; in these instances, the two responses were averaged. The midpoint response for
each scale was entered for items for which participants did not provide a response.
The descriptive statistics for the two task differentiation scales and the role
dissatisfaction measure (all from the "Who Does What?" questionnaire) cannot be
compared to findings from previous research; Cowan, Cowan, Coie, and Coie (1978), the
questionnaire's authors, have not provided descriptive statistics from their instrument.
The mean couple score obtained from this study sample for the household task
differentiation scale was 27.99, within a possible range of 0 to 48. For the child-related
task differentiation scale, the average couple score was 43.63, within possible range of 0
to 84. Traditionally, scores on this scale can range from 0 to 96. However, the decision
was made to exclude three items from the child-related task differentiation scale,