Socioeconomic Variations in the Frequency of Parent Number Talk: A Meta-Analysis

Round 1

Reviewer 1 Report

The manuscript entitled "Socioeconomic variations in the frequency of parent number talk: A meta-analysis" describes the results of a meta-analysis of 12 studies to find the relation between socioeconomic status and parental number talk with young children. The authors report small meta-analytic effects for the association between parental education and number talk (r=.12) and household income and number talk (r=.14). This study does add to the existing literature on caregiver mathematical talk, however, I had several concerns with the manuscript.

Major concerns:

• It would be helpful to have the motivation of the meta-analysis described in more detail in the introduction. What does this specific meta-analytic effect tell us? For instance, if the effect is small, does this suggest that parental math talk is not an area for researchers, educators, and policy makers to intervene on in order to reduce early math disparities?
• My major concern involves the methods. Section 2 (Materials and Methods) needs more information on how the 12 studies were identified. This is the basis of the manuscript, and the estimated effect size is only helpful if readers feel confident that all of the relevant studies were included. Minimally, the authors should include: the timeline for the literature search and/or the years of studies included (i.e., studies from the past ten years, through 2021, etc); the number of unique studies returned by the initial searches; the number of studies screened out; and most importantly - the inclusion criteria for being included in the meta-analysis (i.e., peer-reviewed? in English? had at least one reported association of SES and parent number talk? age of children? etc).
• I question the decision to collapse across experimental conditions... weren't some of the conditions designed to promote parental number talk? That seems to be an entirely different circumstance than control conditions or conditions where parents and children are just asked to interact as they normally would at home.
• I think it would be helpful to add more discussion about the contexts in which these associations were observed - again, it seems pretty different to observe number talk in a setting where researchers have assigned you to a condition and given you math-specific resources to engage with, compared to more naturalistic, home contexts.
• A final thought that could relate to the variability observed in parent number talk is the characteristics of the individual child - for example, their number skills, their interest in math, and their general cognitive abilities (like executive functioning). Parent number talk is dyadic - parents are talking to their children - therefore it is highly likely that children are affecting their parents' talk.

Minor concerns:

• The sentence p. 1 lines 34-36 needs references ("...a domain of parent talk that is robustly predictive of child math skills, but for which the evidence of SES differences is mixed.")
• On p. 2, the reference to Dailey and Bergelson should clarify that the study investigated parent's talk to infants.
• On p. 7, line 204 - I think it is important to be clear that the main variable of interest that will be meta-analyzed is the association between SES and parent number talk.
• In section 2.3 - please report the software edition used to conduct these analyses.
• Did every study contribute an effect size as a correlation (r)? Or were some of these effects converted from means and standard deviations from two groups (i.e., lower/higher SES).
• For several of the moderator analyses, I think it would make more sense to compare the average effect sizes for the two groups rather than report a correlation. For example, comparing the estimated effect sizes for studies conducted in the lab vs. in home - it is harder to interpret a correlation coefficient here compared to seeing the two estimated effects.

Author Response

Please see the response letter we have uploaded.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is original with high scientific quality. I have no comments on improvements and I recommend the paper go forward

Author Response

We sincerely appreciate the reviewer's positive evaluation of our work.

Reviewer 3 Report

The authors used 12 studies in their research, in which they made a meta-analysis correlations between parent number talk, during their interactions with their children at the age 22-79 months. The authors analyzed two aspects of family socioeconomics: which is parent education stage and family income. 

The text uses a wealth of literature on the subject. In terms of methodology, the text was properly prepared. Detailed research results are additionally illustrated with graphical charts.

The text does not require editorial correction.
I accept the text as presented and recommend it go forward.
Congratulations to the authors of interesting research.

Author Response

We sincerely appreciate the reviewer's encouragement and positive feedback on our manuscript.

Round 2

Reviewer 1 Report

Thank you for your thorough responses to my original concerns. I think the manuscript makes a nice contribution to the early math talk literature.