Due to the pandemic, text-messaging interventions have emerged as a promising alternative or supplement to in-person programs. If effective, this modality is not only low-cost but also can be scaled up easily. Since text messaging does not require Internet access, it is also a great way to disseminate information to vulnerable populations.
Over the past few years, I have been involved in designing, adapting, and evaluating positive parenting programs offered via text message. More recently, I have been discussing my results with colleagues interested in implementing similar programs. Since I get asked many of the same questions concerning design and program delivery (e.g., how many messages to send, at what time of day to send them, etc.) that my co-authors and I had when designing our interventions, I thought it would be helpful to summarize here what I have learned. I hope that this blog will also be useful not only to my colleagues at the World Bank but also to those interested in responding to the Early Learning Partnership call for parental interventions (see internal link) and need a starting point for designing how to deliver their interventions.
Too little or too much information?
For many interventions aimed at changing adults’ behavior, including parenting programs, it is easy to assume that more is better. This tendency is even more tempting when one is thinking about offering a light-touch text-message intervention that is cheaper and can be more effective compared to traditional in-person programs. However, raising children can be overwhelming. Factors like financial or time constraints can place added stress on parents. In this sense, by giving participants too much information, it might increase their cognitive demand and cause them to interact less effectively with their children (Cortes et al., 2018).
Considering parents' cognitive load, Cortes et al. (2018) study if text messaging parenting programs can supply too little or too much information. The authors analyze the importance of content and frequency of the text messages in the context of a parenting program (Ready4K) that aimed to improve early literacy through text messages. Participants are parents of four-year-old preschoolers in the Dallas Independent School District (ISD). The authors varied the content and number of messages that were sent to the caregivers and showed that, for this population, a single weekly text was less effective at improving parenting practices than a set of three texts that include both information and encouragement. On the flipside, they also found that five messages per week was too much for parents and did not benefit the children; in fact, more text messages led to higher program attrition and lower parental engagement.
Another danger associated with too much information is that it can cause parents to opt-out of a program, even if the intervention is effective. In a short paper, Fricke et al., (2018) study (among other things) whether more frequent texts and more complex messages cause greater opt-out. After merging individual-level data from eight text messaging programs that foster parent–child interactions and school readiness, they documented two interesting results. First, parents can feel overburdened by programs with too many or too complex texts. For example, programs that sent five texts had a higher dropout rate relative to programs that sent only three messages. Similarly, the dropout rate was higher for programs that sent three texts compared to programs that sent only one message per week. Second, programs that delivered activities that were easy to understand and operationalize, combined with context and encouragement, had lower dropout rates compared to programs that sent impersonal descriptions of activities alone.
Does timing matter?
Another constraint in providing beneficial parenting programs is timing. For example: Are parents more likely to read a text message at 7:30 AM (when they are likely getting the kids ready for school or daycare and themselves ready for work), or at 10:00 AM on the weekend (when they are less likely to be in a hurry)? The time at which messages are sent can greatly determine whether an intervention is beneficial because, even if the program content is effective, it will be less so or not at all if delivered to parents at an inconvenient or impractical hour or on inconvenient day(s) of the week.
Taking into consideration when parenting support is provided, Cortes et al. (2023) compared the effects of a text-messaging program based on whether the text messages were sent to parents either on the weekend or during the week. Their results suggest that, on average, sending text messages on the weekend is more beneficial to children’s outcomes than sending texts on weekdays. Interestingly, they found relevant heterogeneous effects by the child’s baseline skills and the parent’s educational level. The weekday program was more beneficial for higher educated parents and for the initially higher achieving children, whereas the weekend program was more beneficial for less educated parents and for children in the lower half of the baseline skills distribution. The authors argue that these results are consistent with the fact that parenting support works best when parents have time, attention, and need.
To personalize, or not?
Lastly, another relevant question to ask when designing a program delivered by text message is whether the level of personalization is relevant. Doss et al. (2019) noted that the advantage of differentiation (a combination of child-specific information and an increased sense of familiarity) is that it provides parents with specific (and, therefore, potentially more effective) activities that are relevant to their child’s development or characteristics. On the other hand, if the text messages serve as nudges to the caregivers, then personalization would not affect the program's effectiveness but just impose an additional and unnecessary logistic cost.
Doss et al. (2019) also explored the question of personalization by comparing parents who received a generic text-based parenting program and parents who received a personalized version of the same program. The results show that children of parents who received personalized texts had higher reading skills compared to the children of parents who received generic text messages. Parents who received personalized and differentiated messages also reported that it was easier for them to build their children’s reading skills, while parents who received generalized messages did not report this experience.
Text-based parenting programs are a very promising low-cost and scalable approach to improving children’s outcomes. However factors such as content frequency, quality, complexity, and personalization are important to consider to ensure that the intervention is well received, implementable, and effective. Since the text-messaging modality is more salient, it is important to also keep in mind that further evidence is necessary to understand what other types of design factors would increase program engagement and efficacy.
Note: This blog is not meant to be an extensive summary of the evidence. If any of our dear readers have other insights, experiences, or results from rigorous studies to share, please do so.