About this paper symposium

---

Panel information
Panel 27. Solicited Content: Global South

Paper #1
Measuring Parent-Child Interactions in Resource-Constrained Settings: Evidence from an Observational Tool Implemented in Andean Peru
Author information	Role
Dr. Dana Charles McCoy, Ph.D., Harvard Graduate School of Education, United States	Presenting author
Marta Dormal, Harvard Graduate School of Education, United States	Non-presenting author
Kristen Hinckley, Swiss Tropical and Public Health Institute, University of Basel, and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Non-presenting author
Milagros Alvarado, Swiss Tropical and Public Health Institute, University of Basel, and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Non-presenting author
Lena Jäggi, Swiss Tropical and Public Health Institute and University of Basel, Switzerland	Non-presenting author
Abstract
Parent-child interactions are thought to be key drivers of young children’s development around the world, yet tend to be difficult to conceptualize and measure across diverse cultural contexts (Mesman et al., 2018). Although observational approaches have been cited as particularly objective and rigorous means for operationalizing parent-child processes, few observational measures have been validated for use in culturally diverse, low-resource Majority World settings (Bozicevic et al., 2024). The goal of this study is to document the adaptation, implementation, and validation of a 4-minute observational measure of mother-child interactions in Andean Peru. Specifically, we (1) describe the process of adapting an existing measure of mother-child interactions to meet the needs of the local context, (2) establish the factor structure of the resulting measure, and (3) examine the associations between our new measure and existing, mother-reported measures of parenting processes, as well as child development outcomes. Data for the present study came from the endline of a cluster randomized control trial evaluating the impact of a parenting intervention in the Cajamarca region of Peru. Initial analyses have been conducted on a sample of 846 two-year-old children (mean age = 29.15 months; 49% girls) who have completed endline so far. The average number of years of education for mothers in the sample was 9.39 (SD = 4.00). During a visit to participants’ homes, we used an adapted version of the Observation of Mother-Child Interactions (OMCI; Rasheed & Yousafzai, 2015) to measure mothers’ responsiveness, emotional support, and cognitive support while interacting with their children with a picture book. Specifically, we: reduced the original OMCI to include 11 items focused on mothers’ behaviors; adapted the items themselves to align with local conceptualizations of responsiveness and support; and modified the scoring approach to use a tally system. We also measured mothers’ use of stimulation and discipline practices using the Family Care Indicators (FCI; Kariger et al., 2013), and child development outcomes using the directly-assessed Global Scales of Early Development (GSED; McCray et al., 2023). We used confirmatory factor analysis with a tobit estimator to compare three theory-driven factor structures (Muthén, 1989). Results comparing one- and two-factor models that exclude items with low factor loadings are shown in Figure 1, and suggest the two-factor model as the winning model. (A three-factor model failed to converge.) Subscales from the two-factor model (reflecting emotional and cognitive support, respectively) showed small but statistically significant (p < .05) bivariate correlations with stimulation and discipline in expected directions (r = |.05 - .14|). In regression models these subscales also positively and significantly predicted child development above and beyond demographic covariates and mother-reported parenting measures (beta emotion = .08, beta cognition = .17). Collectively, these initial results suggest that the adapted OMCI may be useful for measuring mothers’ cognitively and emotionally supportive behaviors with their children in Andean Peru. The final presentation will include updated results with the full sample, as well as a discussion of implications for the growing number of efforts aimed at measuring, understanding, and supporting parent-child interactions in Majority World countries.

Paper #2
Development of an Instrument to Measure Parenting Cognitions and Practices in High Andean Peru
Author information	Role
Kristen Hinckley, M.Ed., Swiss Tropical and Public Health Institute, University of Basel and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Presenting author
Milagros Alvarado Llatance, Swiss Tropical and Public Health Institute, University of Basel and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Non-presenting author
Dana Charles McCoy, Harvard Graduate School of Education, United States	Non-presenting author
Lena Jäggi, Swiss Tropical and Public Health Institute, University of Basel, Switzerland	Non-presenting author
Marta Dormal, Harvard Graduate School of Education, United States	Non-presenting author
Daniel Mäusezahl, Swiss Tropical and Public Health Institute, University of Basel and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Non-presenting author
Stella Hartinger Pena, Swiss Tropical and Public Health Institute, University of Basel and Universidad Peruana Cayetano Heredia, Switzerland and Peru	Non-presenting author
Andreana Castellanos, Afinidata, United States	Non-presenting author
Maria Luisa Huaylinos Bustamante, Universidad Peruana Cayetano Heredia, Peru	Non-presenting author
Leonel Aguilar, Institute for Computing Platforms, Federal Institute of Technology Zurich, Switzerland	Non-presenting author
Maria Catalina Gastiaburu Cabello, Universidad Peruana Cayetano Heredia, Peru	Non-presenting author
Nerita Gabriela Quispe Roncal, Universidad Peruana Cayetano Heredia, Peru	Non-presenting author
Günther Fink, Swiss Tropical and Public Health Institute, University of Basel, Switzerland	Non-presenting author
Abstract
Despite growing interest in supporting caregivers’ provision of nurturing care globally, few measures of parenting have been validated for use outside of Minority World contexts. It is important to have easy-to-use parent-reported measures that are locally responsive and accessible to caregivers with low levels of formal education. The aim of this study was to develop, describe, and validate a feasible, comprehensive, culturally relevant instrument to measure local parenting cognitions and practices regarding responsive caregiving and early learning within a sample of caregivers of 2- to 5-year-olds in Cajamarca, Peru. We used a multiple methods, phased study design to develop the self-reported parenting instrument, including qualitative interviews, focus groups, cognitive testing, and quantitative pilot testing. The final validation sample included 1,109 mothers of children (mean age = 29.47 months, 50.05% female) from Cajamarca, Peru. The parenting instrument included 38 items with 3-point ordinal response scales, covering parenting beliefs, attitudes, knowledge, and practices regarding responsive caregiving and early learning. We used confirmatory factor analysis (CFA) to examine the instrument factor structure. We assessed concurrent validity with correlations between the instrument and parenting outcomes, including parental stimulation using Family Care Indicators (FCI; Kariger et al., 2013), harsh punishment using Multiple Indicator Cluster Survey (MICS 7; United Nations Children’s Fund, 2024), and learning materials using International Development and Early Learning Assessment (IDELA; Save the Children, 2024), FCI; Kariger et al., 2013), and additional context-specific items. We assessed predictive validity by regressing early childhood development (ECD) on the instrument, using directly-assessed Global Scales for Early Development (GSED; McCray et al., 2023), parent-reported Caregiver Reported Early Development Instruments (CREDI; Waldman et al., 2021), and parent-reported Strengths and Difficulties Questionnaire (SDQ; Youthinmind, 2022). Results of CFAs indicate a two-factor structure reflecting parenting cognitions (α = .81) and parenting practices (α = .79). All standardized factor loadings were greater than .60 (see Figure 1) and model fit was generally adequate (Robust RMSEA = .08, Robust CFI = .89, Scaled SRMR = .06.). Analyses of measurement invariance suggest the instrument performs similarly for caregivers of boys and girls, from high and low socioeconomic status, and living in rural and urban areas. Both parenting cognitions and practices subscales correlated positively and statistically significantly (p < .01) with parental stimulation (rcognitions = .35 and rpractices = .56) and learning materials (rcognitions = .40 and rpractices = .38), and negatively with harsh punishment measures (rcognitions = -.11 and rpractices = -.10). Both parenting practices and cognitions positively and statistically significantly (p < .05) predicted ECD (parent report, β_cognitions= .23 and β_practices= .46, and direct assessment β_cognitions= 1.76 and β_practices= 1.18), and negatively for parent-reported child externalizing problems (β_cognitions= -.11 and β_practices= -.15). Overall, results suggest that this new instrument provides valid and reliable evidence regarding parents’ cognitions and practices related to nurturing care in Andean Peru. This instrument can be used as a starting point for measuring parenting in other Majority World contexts with similar levels of caregiver formal education. Additional practical and policy implications will be discussed in the presentation.

Paper #3
Development, Reliability, and Validity of the Responsive Care Tool for Children 0-3-years-old in Rural Pakistan
Author information	Role
Elizabeth Hentschel, Yale Child Study Center, Yale School of Medicine, United States	Presenting author
Saima Siyal, Development and Research for Children in Early Adolescent Years of Life (DREAM) Non-Governmental Organization, Pakistan	Non-presenting author
Dana Charles McCoy, Harvard Graduate School of Education, United States	Non-presenting author
Henning Tiemeier, Harvard T.H. Chan School of Public Health, United States	Non-presenting author
Aisha Khizar Yousafzai, Harvard T.H. Chan School of Public Health, United States	Non-presenting author
Abstract
Research has shown the importance of responsive caregiving for fostering positive development early in life; however, tools measuring these interactions are often impractical for larger scale intervention trials and in settings with resource constraints. The Nurturing Care Framework (NCF), an influential policy framework developed by the World Health Organization, The World Bank, and UNICEF, has included responsive caregiving as a key indicator that countries should invest in to promote early child development. Despite the policy-focus, no tools currently exist that adequately quantify responsive caregiving with the potential to be implemented in diverse settings (Black et al., 2024; Jeong et al., 2022). Based on the theoretical definition of responsive caregiving, the need for the tool to be relevant for caregivers with children under three years of age, to be freely available, to be usable in both resource constrained settings and with the potential to be used in larger scale intervention trials, we developed an observational tool to measure responsive caregiving (RC tool). The present study provides reliability and validity evidence of the RC tool from Naushahro Feroze, Sindh, Pakistan. Data was collected from 200 randomly selected households, with children under the age of 3, on responsive caregiving, socio-demographic characteristics, early learning, and early child development. The results indicated that the responsive care tool can be feasibly administered in less than 5-minutes in a low-resource setting. Inter-rater reliability was high, with an average Cohen’s kappa coefficient of .95. An exploratory factor analysis found that the tool’s indicators reliably loaded onto two distinct factors, responsive interactions and caregiver-initiated interactions, accounting for 96.01% of the underlying variation in scores. A confirmatory factor analysis reflecting input from modification indices showed satisfactory fit statistics and adequate factor loadings (all above .70). Internal consistencies of the two factors were also high, with alphas of .93 and .83, respectively. Convergent validity of the responsive interactions factor was demonstrated by a strong and positive correlation with measures of psychosocial stimulation, early learning, maternal education, and household wealth. Predictive validity of the responsive interactions factor was demonstrated by a strong and positive association with child development. The caregiver-initiated interactions factor was significantly and negatively associated with psychosocial stimulation and child development. The resulting evidence provides programs with an open access, observational, reliable, and valid measure to quantify responsive caregiving at the program level in low-resource settings. Due to ease of training for assessors, the tool also has the potential to be useful for larger scale intervention evaluations.

Paper #4
Development and Validation of a Father Involvement Measure During Early Childhood in a Resource-Limited Context
Author information	Role
Berta Bartoli, M.Ed., New York University Steinhardt School of Education, Culture, and Human Development, United States	Presenting author
Joshua Jeong, Emory University Rollins School of Public Health, United States	Non-presenting author
Juliet McCann, Emory University Rollins School of Public Health, United States	Non-presenting author
Abstract
Globally, there is mounting evidence about the importance of father involvement for improving child/family wellbeing, reducing gender inequities, and improving population health (e.g., Aguiar & Jennings, 2015; Yargawa & Leonardi-Bee, 2015). In tandem, the involvement of men as caregivers has expanded since the turn of the century (Jeong et al., 2018, 2023) but there are limited tools available for assessing father involvement in low-resource settings. The objective of this study was to develop and validate a scale for measuring fathers’ engagement in caregiving activities for young children and families in rural Kenya. We used baseline data collected in February-March 2023 from a cluster-randomized controlled trial of a parenting intervention aimed at improving early child development in Nyamira and Vihiga counties in Western Kenya. The sample comprised of 460 primary caregivers of children under 18 months of age (91.3% mothers) who were in a relationship with a male caregiver of the young child. The primary caregiver reported on the 25-item father involvement scale. We conducted exploratory and confirmatory factor analyses to determine the dimensionality of the measure and estimated latent variable regression models to examine measurement validity in terms of associations with sociodemographic factors, other paternal and maternal psychosocial measures, and early child development outcomes. The scale had a four-factor structure representing father involvement in (1) childcare activities, (2) play and affection, (3) household chores, and (4) early learning activities (goodness-of-fit-statistics: χ2=340.88, df=269, CFI=0.99, RMSEA=0.03, TLI=0.99, SRMR=0.10; and all factor loadings>0.55). Each domain showed good internal consistency (α=.76-.85; ω=.83-.90) and evidence of construct (i.e., convergent and discriminant) and criterion (i.e., predictive) validity. We found evidence of the measure’s convergent validity by showing that paternal harsh discipline, maternal ratings of fathering quality, maternal-gender equitable attitudes, and couples’ joint decision-making were positively associated with multiple domains of the measure. In addition, we found evidence of the measure’s discriminant validity by showing that maternal depressive symptoms and parenting stress were negatively associated with three domains, childcare activities, play and affection, and household chores. Maternal intimate partner violence (IPV) victimization was also associated with lower paternal involvement in household chores. All domains of father involvement were positively associated with child cognitive, language, motor, and socioemotional development after controlling for sociodemographic factors. Though the magnitude of these associations were small, the early learning domain scores had the largest associations with long-term child outcomes relative to other domains. We provide a valid tool for assessing father involvement in rural Western Kenya and potentially other similar contexts, filling an important gap in the existing literature. Best represented by a four-factor model, our tool provides a more holistic assessment of father involvement compared to prior survey measures that have narrowly quantified father involvement in terms of one singular dimension of caregiving behavior. Considering the wide range of positive associations with maternal, family, and ECD outcomes, our results reinforce how father engagement should be holistically conceptualized in terms of the supportive activities that fathers can engage in for both their partners and young children.

⇦ Back to session