Integrating qualitative and quantitative methods to develop a comprehensive coding manual: Measuring attention to mathematics in play contexts

This coding manual was created to quantify and characterize the ways in which parents and their preschool children attend to mathematical features during play. We operationalized attention to mathematical features as instances during which parents or children made verbal references to quantitative (i.e., number, quantity, part-whole relations) or spatial (e.g., size, shape, spatial location, positional orientation) features. Through analysis of videotapes and transcripts of parent-child play sessions, we coded these verbal communications as questions, responses, mentions, or repetitions; and coded corresponding nonverbal support as either explicit, implicit, or irrelevant to the mathematical features coded. Analyses of independent coding by two coders indicated that we reliably coded instances of attending to mathematical features and the ways in which dyads talked about these features. Codes for nonverbal behaviors were less reliable and may require further development to improve reliability. The reliability analyses on verbal communication also illustrated the Kappa paradox—in our case, the high proportion of feature mentions led to a low value of Cohen's Kappa even when the percent agreement and Gwet's Agreement Coefficient 1 were high. The coding scheme served as a framework for our recent report on the frequency of parents' and children's attention to mathematical features and the association between parents' and children's attention to these features (Chan et al., 2020). This coding manual leverages the benefits of a mixed-methods approach by coding qualitative observations for quantitative analysis, and may provide other researchers with a framework for coding attention to mathematical features in a play context and developing manuals for analyzing other aspects of parent-child interaction.