Participants

Ruby, which variable we will include in this section? Please notice that results has informations that might be moved into this section.

Procedure

Measures

For this study, the participants rated their aspects of Home Language and Literacy through standardized questionnaires, in addition to specific questions about the availability of reading materials and the mother’s frequency of reading to the child.

The Child interest in litteracy refers to child engagement in literacy activities. This scale is composed of 10 items and acess the frequency a child independently engage in activities such as looking at books and telling stories. Previous study has fond a Cronbach’s α of .81.

Adult’s Value placed on literacy (Ruby, could you please describe this scale?)

Press for achievement refers to the frequency mothers taught their children early academic skills such as letter sounds, the alphabet, etc. This scale is composed of 6 items. Previous study has found a Cronbach’s α of .80.

Reading with child refers to the frequency of mother and father read with child. This scale was answered on an ordered 7-point scale: (1) never; (2) sometimes a year, (3) once a month, (4) 2-3 times a month, (5) once a week, (6) 2-4 times a week, and (7) 5-7 times a week. To further investigate the effect of reading, a dose category was created using 1 SD above the mean to define high dose.

Attitude towards reading address questions about enjoyment and dissatisfaction of child-caregiver shared-reading activities. The scale is formed of 13 items. Previous study has found Cronbach’s α of .81).

Two questions were used to check the availability of reading materials: ¿Cuántos libros para adultos hay en tu casa aproximadamente? and ¿Cuántos libros propios tiene tu niño/a? (Por favor no incluyas libros de hermanos mayores). The scale was transformed to 1 to “Ninguno / 0 libros”, 2 to “1 to 5 books”, 3 to “6 to 10 books”, 4 to “11 to 20 books”, 5 to “21 to 30 books”, 6 to “31 to 40 books”, 7 to “41 to 50 books”, and 8 to “more than 50 books”.

Data Analysis

Successive steps were taken to analyze the data of this research. The categorical variables were analyzed through percentages and frequencies. The Chi-squared test was used to check whether the variables were associated. The items of the five different specific questionnaires used in this research were checked in terms of its psychometric properties. It’s recognized that the Principal Component Analysis (PCA) is a formative latent variable approach useful for i. the classification of the data and ii. to reduces its dimensionality by checking its set of patterns and finding a new set of uncorrelated variables (Components) that can retain the most of the information. This approach assumes that the extracted components are dependent on the indicators (items) measured with little or no error. In other words, indicators are causes of an underlying latent variable, rather than its effects.

There are many rules of thumb available to help choose the number of components to extract. This study relied on the result of the Parallel Analysis using a Monte Carlo approach with 2000 randomly simulated datasets. The solution obtained was acceptable if i each factor had three items or more, ii with 0.3 as the minimum of its factor loading, and iii the overall solution was theoretically meaningful and interpretable. The data analysis was conducted using R 3.5.1.

1 - Child interest in litteracy

The Child interest in litteracy refers to child engagement in literacy activities. This scale is composed of 10 items: “a. Mirar un libro solo o sola”, “b. Mirar una revista”, “c. Mirar un periOdico”, “d. Escuchar musica”, “e. Colorear solo o sola” , “f. Pretender escribir”, “g. Relatar un cuento”, “h. Cantar canciones”, and “i. Crear una tarjeta de felicitaciones para alguien”.

The parallel analysis has suggested two as the optimal number of components (Figure x). This solution accounts for 0.47 of the variance on the data. The Cronbach’s Alpha was 0.73 and the Omega’s Mcdonald was 0.75. These results ensure the reliability of the data.

## Parallel analysis suggests that the number of factors =  NA  and the number of components =  2

2 - Caregiver litteracy

The adult’s value placed on literacy adult’s engagement in literacy activities. This scale is composed of 12 items: a. Leer un libro o una novela, b. Leer la Biblia o textos religiosos, c. Leer una revista, d. Leer un periódico, e. Leer los anuncios comerciales, f. Usar un diccionario, g. Usar recetas de cocina, h. Hacer una lista (por ejemplo, para el supermercado), i. Dejar un recado escrito para alguien, j. Mandar un mensaje (por correo o por correo electrónico), k. Mandar un mensaje de texto por celular, and l. Leer algo por el Internet.

## Parallel analysis suggests that the number of factors =  NA  and the number of components =  2

The first dimension was composed of 5 items and its Cronbach’s alpha was 0.84, whereas the second dimension had 0.75 as its Coefficient. The Omega’s Mcdonalds was 0.82 and 0.75 respectively. The Cronbach’s Alpha for all items was 0.84 and Omega’s Mcdonald was 0.86.

3 - Press for achievement

The Press for achievement is composed of 6 items: “a. Hablar con tu ni?o/a del alfabeto”, “b. Hablar con tu ni?o/a sobre los sonidos que corresponden a las letras”, “e. Ayudar a tu ni?o/a a seguir instrucciones impresas de un juego o un juguete”, “g. Ayudar a tu ni?o/a a escribir”, " i. Llevar a tu ni?o/a a una librer?a“,”j. Leer letreros con tu ni?o/a". The parallel analysis has revealed that one dimension can describe the data and 0.45 of the variance can be attributed to this component.The Cronbach’s Alpha was 0.75 and Omega’s Mcdonald was 0.74, what ensures the reliability of the data.

## Parallel analysis suggests that the number of factors =  NA  and the number of components =  1

4 - Attitude

The Attitude scale is composed of 13 items: “a. A mi ni?o/a le gusta que le lean”, “b. Me siento cercano/a a mi ni?o/a cuando leemos”, “c. Tengo que castigar o disciplinar a mi niño/a cuando tratamos de leer”, “d. Quiero que mi ni?o/a valore los libros”, “e. No le leo a mi ni?o/a porque no se queda quieto/a”, " f. Me parece aburrido o dif?cil el leerle a mi ni?o/a“,”g. Cuando leemos trato de sonar entusiasmada para que mi ni?o/a est? interesado/a“,”h. A?n si quisiera, estoy muy ocupado/a y muy cansado/a para leerle a mi ni?o/a“,”i. No le leo a mi ni?o/a porque no tenemos nada que leer“,”j. No le leo a mi ni?o/a porque no hay espacio y no hay un lugar tranquilo en la casa“,”k. No le leo a mi ni?o/a porque tengo otras cosas m?s importantes que hacer como padre/madre“,”l. Le leo a mi ni?o/a cuando ?l/ella quiere“,”m. No le leo a mi ni?o/a porque no s? leer o tengo dificultades para leer". The parallel analysis has suggest to retain two dimensions that account for 0.51 of the variance.

## Parallel analysis suggests that the number of factors =  NA  and the number of components =  2

The first dimension was composed of 8 items and its Cronbach’s alpha was 0.85, whereas the second dimension had 0.85 as its Coefficient. The Omega’s Mcdonalds was 0.86 and 0.74 respectively. The Cronbach’s Alpha for all items was 0.86 and Omega’s Mcdonald was 0.74.

5 - Availability

A scale having two items only has been recognized as problematic and the methods for accessing its reliability are not free from academic dispute or criticism. Recent literature suggest that the most appropriate reliability statistic is the Spearman-Brown coefficient and the standardized coefficient alpha (Eisinga, Grotenhuis, and Pelzer 2012). Both computations were explored. The Standardizes alpha was 0.72 and the Spearman-Brown adjustment was 0.72. Along with these results, the next table reports descriptive information.

6 - Reading

Reading with child refers to the frequency of mother and father read to the child. Because this scale is also composed of two items, its reliability was acesses by Standardized alpja (0.56), and the Spearman-Brown adjustment 0.72. Along with these results, the next table reports descriptive information.

In line with previous research, the average of the scale items was computed to work as an index of this behavior.

Child interest

The first construct explored was the Child's interest in literacy. Country (F(3,242) = 5.81, p < .01) and income (F(4,242) - 2.79, p =.03) affected child’s interest in literacy.

## 
## 
## ANOVA results using mean_child_interest as the dependent variable
##  
## 
##              Predictor      SS  df      MS       F    p partial_eta2
##            (Intercept) 1612.86   1 1612.86 2277.99 .000             
##        factor(Country)   17.14   4    4.29    6.05 .000          .05
##  factor(mother_school)    2.06   1    2.06    2.91 .089          .01
##  factor(father_school)    2.28   1    2.28    3.22 .074          .01
##      factor(bilingual)    0.61   1    0.61    0.86 .355          .00
##         factor(salary)    0.51   1    0.51    0.71 .399          .00
##                  Error  294.53 416    0.71                          
##  CI_90_partial_eta2
##                    
##          [.02, .09]
##          [.00, .03]
##          [.00, .03]
##          [.00, .02]
##          [.00, .01]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

The post hoc comparison revealed a significant effect for Ecuador and Guatemala.
Ruby, we shoud always look at comparisons column. When the estimate is negative, it means the first group has lower scores than the second one. I see some patterns in the results, but I my text just describe it lineary. It will be much more elegant to have an overall text describing these results.

The comparison between Mexico and Ecuador was significant (d = -0.427, p-value adjusted = 0.00084), so do for the comparison between the USA and Ecuador (d = -0.571, p-value adjusted = 0.00094). These results show that the Child’s interest in literacy is lower in Mexico and in the USA than Ecuador. The same pattern was found in Guatemala. The results show that Mexico and the USA participants in average have lower results on Child’s interest in literacy: d = -0.362, p-value adjusted = 0.0056, and d = -0.507, p-value adjusted = 0.0041 respectively.

The comparison between all levels of income was also carried out. Only a marginal effect was found in the comparison between Double the minimum salary and Between extreme poverty and below minimum salary, whereas the participants with higher earnings had lower average on this child’s interest in literacy (d = -0.39, p-value adjusted = .08). Ruby, we never know if people answered this question properly..

Adult´s value paced on lieracy

An ANOVA did not find any significant main effect between predictors explored on the Adult´s value placed on literacy: Country (F(3,221) = .790, p = .54), Mother Education (F(7,221) = 1.61, p = .13), Father education (F(7, 221) = .714, p = .66), Bilingualism (F(1,221) = 1.17, p = .28), and income (F(4,221) = .434, p = .78).

## 
## 
## ANOVA results using mean_adult_literacy as the dependent variable
##  
## 
##              Predictor     SS  df     MS      F    p partial_eta2
##            (Intercept) 584.57   1 584.57 399.59 .000             
##        factor(Country)  22.89   4   5.72   3.91 .004          .04
##  factor(mother_school)  16.85   1  16.85  11.52 .001          .03
##  factor(father_school)   1.87   1   1.87   1.28 .259          .00
##      factor(bilingual)   5.82   1   5.82   3.98 .047          .01
##         factor(salary)   0.72   1   0.72   0.49 .484          .00
##                  Error 542.74 371   1.46                         
##  CI_90_partial_eta2
##                    
##          [.01, .07]
##          [.01, .06]
##          [.00, .02]
##          [.00, .03]
##          [.00, .01]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

Press for achievement

Ruby, should we check the marginal effect of bilingualism?

## 
## 
## ANOVA results using mean_press_achievement as the dependent variable
##  
## 
##              Predictor      SS  df      MS       F    p partial_eta2
##            (Intercept) 1423.49   1 1423.49 1188.07 .000             
##        factor(Country)    6.73   4    1.68    1.40 .232          .01
##  factor(mother_school)    1.64   1    1.64    1.37 .243          .00
##  factor(father_school)    0.00   1    0.00    0.00 .963          .00
##      factor(bilingual)    0.80   1    0.80    0.67 .414          .00
##         factor(salary)    0.03   1    0.03    0.03 .868          .00
##                  Error  506.82 423    1.20                          
##  CI_90_partial_eta2
##                    
##          [.00, .03]
##          [.00, .02]
##         [.00, 1.00]
##          [.00, .01]
##          [.00, .00]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

Availability

Ruby, should we check the marginal effect of bilingualism?

## 
## 
## ANOVA results using mean_availability as the dependent variable
##  
## 
##              Predictor     SS  df    MS     F    p partial_eta2
##            (Intercept)  60.90   1 60.90 35.02 .000             
##        factor(Country) 122.69   4 30.67 17.64 .000          .15
##  factor(mother_school)  28.93   1 28.93 16.64 .000          .04
##  factor(father_school)   6.77   1  6.77  3.89 .049          .01
##      factor(bilingual)   4.81   1  4.81  2.77 .097          .01
##         factor(salary)  29.84   1 29.84 17.16 .000          .04
##                  Error 693.94 399  1.74                        
##  CI_90_partial_eta2
##                    
##          [.09, .20]
##          [.01, .08]
##          [.00, .03]
##          [.00, .03]
##          [.02, .08]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

Reading

Ruby, should we check the marginal effect of bilingualism?

## 
## 
## ANOVA results using mean_reading as the dependent variable
##  
## 
##              Predictor      SS  df      MS      F    p partial_eta2
##            (Intercept) 1347.40   1 1347.40 715.06 .000             
##        factor(Country)   12.01   4    3.00   1.59 .175          .02
##  factor(mother_school)    1.16   1    1.16   0.61 .434          .00
##  factor(father_school)    0.31   1    0.31   0.17 .684          .00
##      factor(bilingual)   14.29   1   14.29   7.58 .006          .02
##         factor(salary)    0.79   1    0.79   0.42 .517          .00
##                  Error  744.31 395    1.88                         
##  CI_90_partial_eta2
##                    
##          [.00, .03]
##          [.00, .01]
##          [.00, .01]
##          [.00, .05]
##          [.00, .01]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

Attitude

## 
## 
## ANOVA results using mean_attitude as the dependent variable
##  
## 
##              Predictor     SS  df     MS       F    p partial_eta2
##            (Intercept) 501.17   1 501.17 3115.91 .000             
##        factor(Country)   1.04   4   0.26    1.61 .171          .02
##  factor(mother_school)   0.36   1   0.36    2.22 .137          .01
##  factor(father_school)   0.18   1   0.18    1.14 .287          .00
##      factor(bilingual)   0.02   1   0.02    0.10 .748          .00
##         factor(salary)   0.79   1   0.79    4.91 .027          .01
##                  Error  55.81 347   0.16                          
##  CI_90_partial_eta2
##                    
##          [.00, .04]
##          [.00, .03]
##          [.00, .02]
##          [.00, .01]
##          [.00, .04]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared

1 - Child interest in litteracy

Regarding Child interest in litteracy, 9 items remained from the previous PCA carried out. The following table reports the proportions of each cattegory.

Its also possible to consider each variable as a continuous variable and compute the descriptive statistics for each one.

2- Caregiver literacy

Regarding Child interest in litteracy, this construct refers to caregivers’ reports about their own involvement in literacy-related activities. 12 items remained from the previous PCA carried out.

The following table reports its descriptives.

3 - Press for achievement

Refers to caregiver’s report about the frequency in which they teach their children early print academic skills such as the alphabet and letter-sound correspondence.

4 - Press for Attitude.

Please take into consideration that items Q18c_punish , Q18e_dific, Q18f_boring, Q18h_busy, Q18i_nothing, Q18j_space, Q18k_imp, Q18m_noread will be coded into the oppositve direction.

5 - Availability

Refers to amount of child and adult owned reading material, and findings are mixed

6 - Reading

Refers to the self-reported frequency in which specific caregivers engage in shared-reading with their child (e.g., mother, father).