The Influence of Home Language and Literacy Environment and Parental Self-Efficacy on Chilean Preschoolers’ Early Literacy Outcomes

Reviewer 1

Response to reviewers

Comment 1: This paper addresses an interesting and important topic. Overall, the manuscript is well written and clear, and the method and analyses are appropriate to the topic of study. Below are some minor points that can be addressed:

Thank you for reviewing our manuscripts and for providing very valuable feedback. We have tried to address all your comments and questions in the following section. All changes are highlighted in the revised manuscript.

Comment 2: Introduction:

Move current study paragraph to before method, to improve flow from introduction to method sections.

The paragraph has been moved and is now the introductory paragraph of the Methods section (p. 8-9).

Comment 3: Method:

Report Cronbachs alphas for the self-efficacy questionnaire and include a little more information about how the items and subscales were decided upon and arrived at.

In the ‘Narrative skills’ section in the method the font size and spacing seems to change and a number of sentences are repeated.

Thank you for reminding us of these measures. They are now reported in the Methods section (see p. 11).  As Mendive et al. (2020) point out, the HLLE eight items used in the current study derive from Romero-Contreras (2006) questionnaire and it asks about formal and informal activities related to reading as well as attitudes that promote interest and engagement in reading and language. We have included more information about the items and subscales on pg. 11.

Thank you for noticing the font size change, we have also made the changes in that section.

Comment 4: Results:

In section (b) self efficacy in the results section it is not clear why dimension 2 was not included in the analysis

In section (c) there is a mismatch between the research question (alphabet knowledge), and the second sentence which refers to narrative skills. There is not a similar sentence for alphabet knowledge.

The text in the boxes for models 1 and 2 is unclear and not possible to read

We appreciate your comments. We have included an explanation as to why dimension 2 was excluded from the analysis. The items of dimension 2 refer to parental perception of social support within the school, and do not tap into self-efficacy. Psychometric analyses also led us to decide to eliminate dimension 3 because it asks for frequency of activities. Thus, the analysis pertain dimension 2 only.

We apologize for the typo, we had left out narrative skills. That is now included in the question for section b, which now reads: Is HLLE positively and significantly associated with higher levels of alphabet knowledge and narrative skills? We have also added a sentence that describes alphabet knowledge growth. These explanations help readers understand the answer to this research question which comes in the paragraph on p. 19, below table 5.

Thank you for noticing this. We have redone the text boxes for models 1 and 2.

Comment 5: Discussion:

The position of the limitations section after the conclusion is unusual, is this a requirement of the journal style?

Thank you for noticing this.  In line with other articles in Education Sciences, we have moved the limitations section so that they now appear on p.23, before the conclusions.

Reviewer 2

Response to reviewer 2

Comment 6: Thank you for the opportunity to review this paper. I enjoyed reading it and I believe it is an important topic and can be a valuable contribution to the field. I appreciate that it is conducted with low-SES families in Chile which is very important to know about literacy outcomes and HLLE in different cultures and countries and how different interventions may have an impact on them. I do believe the paper could be strengthened and needs further clarification, especially in the methods and results section.

Thank you so much for your insights and comments. Your feedback is very valuable in improving our manuscript, and we have tried to address all of them.

Comment 7: Introduction:

Overall, the introduction was clear and easy to read, but I do have two minor comments

1. Line 48- clarify what “these two skills” refers to. It was unclear.
2. Lines 57-60. This is a great point and a great justification for why this research is so important and relevant. It would strengthen your argument if you could expand this by adding a sentence or 2 about how delays in reading have long term impacts not just for academic performance when kids are young but even later in school and even after they leave school.

1. Thank you for your comments. We have changed the wording on line 48 to clarify which skills the sentence refers to.
2. We agree with you about the importance of stressing the consequences of reading delays and have added two sentences to strengthen the argument. These are highlighted in yellow.

Comment 8: Methods:

The authors did a good job organizing this section.

1. HLLE survey- the Romero-Contreras 2006 citation is not in the reference section.

Thank you. We have added the reference: Romero-Contreras, S. (2006). Measuring language and literacy-related practices in low-SES Costa Rican families: Research instruments and results. Doctoral dissertation, Harvard University, Cambridge, MA

Comment 9: Methods:

4. It would be interesting to see the pre and post means, SDs, Ns, p-values, and effect sizes for the 7 items of the HLLE in a table in a supplementary table. Only 3 of these items are referenced as being significant and not until the discussion.

We have included a supplementary table in the Appendix with the information you mention. The table is on p. 39.

Comment 10: Methods:

5. Self-efficacy- more detail is needed regarding this measure, in general.

We have added a sentence on p. 11 (highlighted) that explains how the survey was designed by the authors following Sanders & Wooley (2005) who report parenting practices best predicted by task-specific measures of parental self-efficacy. We hope that this helps better understand the survey’s purpose and use.

Comment 11: Methods:

6. Was this measure developed by the authors or was this a preexisting survey (if so, it should be cited)? If it was developed by the authors, more information of how it was developed would be important to include, such as why the specific Dimensions and items were chosen.

This measure was developed by the authors (MV and CdlM) at Fundación Alma with the purpose of measuring parental self-efficacy in the context of Alma’s shared reading intervention. We explain this in the previously mentioned comment and further explain why one of the dimensions of the survey was not used in this particular study (pages 11 and 12). We hope these additions are helpful. Psychometric information about the measure is included in the Appendix.

Comment 12: Methods:

7. Relatedly, information should be included about whether or not any psychometric statistics such as factor analysis was conducted to see if the items fell into these dimensions. Alphas for internal consistency for each dimension should also be included. Finally, do these dimensions all go together? It would seem like Dimension 2 does not fit with the other two dimensions as well, but it would be good to see if this is true statistically, or why it was included as a dimension (and it's 2 items) were included in this measure – which is why I think adding more context and information will be very important for the readers, as noted in my previous comment.

Thank you for suggesting the inclusion of more psychometric information for this assessment. We have included alphas for each dimension on p. 13. Alphas for preparedness (dimension 1) were higher than those for belonging and frequency. Factor analyses were also conducted and we have included them as Appendix 2 so as not to go over the word count of the manuscript.

Comment 13: Methods:

8. My main concern with this measure overall is that I’m weary about referring to it as a “parental self-efficacy” survey because I do not believe it is truly measuring self-efficacy, at least not in the way it is described in the literature in the traditional sense or in the way the authors seem to have operationalized it in the literature review. From what I can tell from the description provided about the measure, Dimensions 2 and 3 do not seem to measure self-efficacy. Dimension 1, I could see it as self-efficacy, but it is a very specific self-efficacy related to only home language and literacy activities as opposed to other measures of parental measures of more global parent self-efficacy described in the literature and more frequently used in other studies.

We agree with the reviewer that dimension 1 is the one that best refers to self-efficacy and it is strictly related to home language and literacy activities rather than a more “global” measure of this construct.  The measure was designed by the intervention team to examine compared perception of three dimensions that would be impacted by the intervention: self-efficacy, social support at school, and frequency of language and literacy activities at home. This explains why the two items in dimension 2 were excluded from the analysis, since they refer to parental perception of social support at school. Dimension 2 taps into how parents perceive themselves as efficacious within the context of other parents and of school staff. Dimension 3 asks for frequency of activities and thus we did not use it as part of the self-efficacy measure. To provide more evidence of the validity of the survey, we have added some psychometric statistics for the survey, as explained in the previous comment. We provide factor analysis results to show how items perform in relation to two factors.

Comment 14: Methods

9. ALMA Family Literacy Program- it would be helpful to have more information about the Alma project and if it was designed specifically to be implemented for Chilean families (or Latin American families) or if it was adapted from a program designed in the US for English speaking families. This is relevant given the cultural differences in the ways that white middles class families in the US on which most literacy research is based, and Latin American families read.

Thank you for this suggestion. We agree that more information will facilitate the readers’ understanding of the program. We have added a couple of sentences on p. 14 to explain this in more detail. The program was specifically designed for Chilean families and has not yet been implemented in other Latin American contexts that we know of. However, we have presented the program’s results in international conferences. We perceive an interest among researchers of other Latin American countries to pilot the program, but it would be necessary to make the contextual, cultural, and linguistic adaptations needed.

Comment 15: Results:

1. There was no justification in the Introduction or Methods of why the results were analyzed by grade level. I think this was interesting, but it was not mentioned prior to the Results section.

We talk about results by grade level on p. 17 and beginning of p. 18. We report ANOVA results by grade levels for the children’s measures. We have added two sentences in the Analyses subsection of the Methods section, p. 15. We also analyze theself-efficacy survey rsults on p. 18, where we say that “Perceptions of Kindergarten children were lower than those of pre-Kindergarten parents for both sub scores; however, these differences were not statistically significant.” For HLLE, on p.19 we mention that “HLLE scores were higher among the parents of kindergarten students, compared to parents of PK students”

Comment 16: Results

11. Lines 347-351: add statistics p, and effect sizes for both alphabet knowledge and narrative skills. Effect sizes will be particularly useful to see the magnitude of the effect

We have redone the tables to include p statistics, effect sizes, and ranges. We have also reorganized the descriptive statistics results explanation on p. 17.

Comment 17: Results

12. Is the paragraph starting on line 352 referring to the comparison across grade levels? If so, this should be clarified because as of now it is not clear if the first sentence refers to the overall analysis or the ones by grade level. Additionally, it would be good to include the scores or gains in text so the reader can see the numbers without having to refer to the tables and to avoid any further confusion.

Thanks for pointing to this. The paragraph refers to the comparison across treatment and control group. The analysis by grade level begins on line 359, and we believe this should be a new paragraph, so we have made the change for a new paragraph to begin on line 359. We hope this is helpful in terms of clarification. We have also added the scores in the text.

Comment 18: Results

13. Line 377: Why was dimension 2 not included in the analyses?

We have added an explanation about the decision to exclude dimension 2 on pp. 11 and 12. See also the explanation to comment 10 above.

Comment 19: Results

14. Please clarify what statistical analyses were conducted for self-efficacy and include the appropriate statistical test results in text and in table 4.

We have redone the table to include statistical test results for each item in dimension 1 of the self-efficacy measure (preparedness), including differences, p values, Cohen’s d and number of participants by grade level. The table appears on p. 19.

Comment 20: Results

15. If possible, add effect sizes to all tables

Thank you for this feedback, we have redone the tables to display effect sizes.

Comment 21: Results

16. Tables 3-5 were challenging to read and should be clearly labeled with headings indicating M(SD) and Range like the previous tables. It would also be helpful to have the Ns if the number of pre-K and Kindergarten children in the treatment and control groups are different.

We have redone tables 3-5 so that information is more clearly labelled and includes the data you suggested.

Comment 22: Results

17. The figures for Models 1 and 2 were hard to read. The boxes need to be clearly labeled with phrases or short descriptions readers can know what each box represents.

We have made changes so that the models are more readable now. We hope these changes facilitate understanding.

Comment 23: Results

18. Line 457-458: add p-value for “did not reach significance”

 We have added the p-values on p. 24.

Comment 24: Discussion:

19.  The authors conducted analysis by grade level, but these results are not discussed. I would like to hear more about your thoughts about these grade level differences for some of the outcomes but not others.

We have added comments by grade level throughout to provide a clearer picture of the differences between PK and K students. We have used effect sizes and ANOVA results to talk about these differences in the results section.

Comment 25: Discussion

20.  There is research on Hispanic/Latin American families' reading beliefs and practices and literacy interventions that I think should be included in the discussion to strengthen your findings and set your study in context, as well as acknowledge the importance of cultural differences in the way literacy is viewed and approached by parents and why it is important for literacy programs to be culturally sensitive.

We have added some references of studies conducted in Latin America in the discussion and we have included a brief discussion of findings that are in line with those found in the current study.

Comment 26:

Discussion

21.  Do you think these results are unique to families in Chile or are they generalizable to other countries in Latin America?

This is a great question, and we thank you for bringing it up. We have evidence that points that literacy practices among Chilean families vary considerably (see, for example, Mendive et al., 2020) where the HLLE was used to show great variability in literacy practices of Chilean families within the same SES group. A previous study (n=188 families across different SES), showed that there were similarities in the approximations to literacy observed in families of different SES levels in more developed countries and those in Chile.  Parents with higher levels of education engage in shared reading with their children earlier and more often than those with less education. High SES parents focus more on alphabet instruction than those in low SES contexts. Similarly, parents from lower SES homes engage in conversations with their children to control their behavior or give orders, whereas parents from higher SES contexts use language to engage in conversations or explanations (Strasser & Lissi, 2009). In line with this a national study of second grade reading predictors showed that alphabet knowledge, word reading, and narrative skills were the components that best explained second-grade reading scores (Agencia de Calidad de la Educación, 2018). These findings highlight the need to promote literacy environments that contribute to learning, especially in contexts where social support and practices can compensate for risk factors associated with poverty. We have included this in the discussion section (p. 25-26).

Comment 27: Discussion

22.  Did you have any information about the number of completed workshops the parents attended? Maybe parents who were more engaged with the workshops had kids with highest gains and also higher HLLE. It could be an interesting future direction to mention or just something you might have thought about already.

We do. We did not include that analysis in this manuscript but would be willing to do so. While we believe it might be interesting to readers, we chose to leave it out because it would extend our manuscript beyond the word count.