Introduction

There is strong evidence showing that emotion-focused parenting programs are effective in improving child emotion regulation and reducing mental health problems [1,2]. However, traditional barriers to parenting programs, including in-person delivery, specific locations, and set times [3,4], mean that only a third of parents who are offered parenting programs enrol in them, and up to 60% of parents drop out [5]. The last two decades have seen a substantial shift towards online programs [6], but have not been associated with improvements in retention and engagement [7], perhaps because they have not been specifically adapted for the online context [8].

Online ecological momentary interventions offer a way to improve engagement by delivering tailored support in the moment when it is most needed [9]. Although these interventions have demonstrated benefits in reach and outcomes in adult mental and public health contexts [10], they have not yet been utilised for parenting programs. To inform the development of a parenting program as part of larger project, this study sought to identify a brief set of items from pre-existing measures suitable for measuring in-the-moment parent and child emotion dysregulation and negative affect. This study also sought to determine when parents were most likely to engage with smartphone app prompts, and which difficult parenting situations and times of day were associated with higher levels of parent and child dysregulation and engagement.

Emotion regulation refers to the ability of individuals to understand emotions and effectively manage the expression of emotions in a socially acceptable manner [11]. A moderate-to-strong evidence base shows that poor emotion regulation skills (i.e., high levels of emotion dysregulation) in childhood are associated with later adverse outcomes in domains such as academic achievement [12], peer relationships [13,14], and behaviour difficulties [15]. Patterns of high emotion dysregulation over childhood are also associated with poor mental health and suicidality in subsequent years [16–18]. Indeed, strong evidence from a meta-analysis (35 studies, combined N = 12,516) supports a moderate to large association between adaptive (mean effect size = 0.29–0.51) and maladaptive (mean effect size = 0.21–0.55) emotion regulation strategies and depressive and anxiety symptoms in children aged 11−18 years [19]. Another meta-analysis, conducted by Compas et al [20], showed a moderate negative association between emotion regulation and internalising and externalising behaviours in children aged 5−10 years (19 studies, mean effect size = 0.30–0.32) and a small to moderate negative association in adolescents aged 11−19 years (14 studies, mean effect size = 0.13–0.20). Existing evidence therefore suggests that equipping children and young people with strong emotion regulation skills in early childhood may be an effective means of reducing population-level mental health problems.

Emotion socialisation theory proposes that parenting beliefs and practices, and the family emotional climate, directly affect children’s emotional development [21]. This includes the way in which parents understand and regulate their own emotions, the way in which they display emotions, and the way in which they communicate about emotions with their children [22]. In turn, poor parent emotion regulation has been found to be associated with poor child emotion regulation [23,24], leading to calls for programs that focus on improving both parent and child emotion regulation skills [25]. Parents have a particularly strong influence on child development in early childhood, a time during which children are almost solely reliant on their primary caregivers, and consequently play a significant role in shaping children’s developing emotion regulation skills [26].

Parenting programs have demonstrated efficacy in improving child developmental outcomes, including emotion regulation skills [27]. England-Mason and Gonzalez [28] conducted a review of parenting programs anchored in emotion socialisation theory and measuring change in emotion regulation skills of children aged 0–6 years. Of an initial 1,117 articles retrieved for screening, 12 articles identified three distinct intervention frameworks developed for parents of young children, with two progressing past the pilot stage: Tuning in to Kids and Parent-Child Interaction Therapy-Emotion Development. Both programs showed benefits in improving parents’ coaching of their children’s emotions [28]. Of the two programs, Tuning in to Kids has the strongest evidence for enhancing emotion regulation in parents and children. The program is a two-hour, face-to-face program delivered over a six-week period by a trained facilitator in a group setting, and has been adapted to suit children at different ages (Tuning in to Toddlers, 1–3 years; Tuning in to Kids, 4–10 years; Tuning in to Teenagers, 11–17 years), as well as a program designed specifically for fathers (Fathers Tuning in to Kids), with all variants found to improve parent emotion socialisation [29–31].

The second program, the Parent-Child Interaction Therapy-Emotion Development intervention, posits that depression in young children is caused by poor emotion regulation [32]. A randomised controlled trial comparing an intervention group to a waitlist control group found that, after 20 sessions with a trained facilitator over an 18-week period, parents showed higher levels of emotion-focused parenting techniques after receiving the intervention [33].

Despite evidence of program efficacy, existing parenting programs have had low uptake and high attrition in the general parenting population [4,34,35]. The aforementioned studies, Tuning in to Kids and Parent-Child Interaction Therapy-Emotion Development, reported a 17–20% and 30% attrition rate, respectively [32,36]. This low rate can be partly attributed to barriers to access and perceived burden. Butler et al [3] conducted a systematic review of qualitative studies that analysed parent experiences of parenting programs (26 studies, combined N = 822), and found that barriers to parent engagement and retention were often centred around difficulties in physically attending sessions and difficulties in scheduling time to partake in programs. Thus, although programs like Tuning in to Kids result in improvements to parent emotion regulation skills, the original in-person group format relies on multi-week, face-to-face sessions with trained facilitators. For many parents, this requires them to travel, arrange childcare, and attend at specific times, factors often cited as barriers to engagement [37].

The last two decades have seen a significant shift towards online programs [6], particularly for use with smartphones, however, this has failed to significantly transform rates of retention and engagement, with short-term follow up attrition rates 24% and long-term follow up rates 36%. [7]. Online parenting programs generally offer a modified version of an existing program adapted for an online space, although they still require parents to devote significant time to them, a key barrier to program engagement. As such, even efficacious programs delivered online have high rates of attrition and disengagement [38–40]. Perhaps in part due to these barriers, parents often seek out non-evidence-based online support for immediate answers to their parenting difficulties, making use of parenting forums and social media to ask other parents for advice [41–43], even if parents themselves show scepticism towards some content found online [44]. The results of a qualitative study by Moon et al [45] found that mothers who used the internet for parenting advice felt the amount of advice overwhelming, evidence-based or otherwise. Regardless, parents expressed that a key aspect of online advice was that they were able to find information specific to them. Mothers appreciated the timeliness and relevance of mobile phone apps, allowing access to information when and how they needed it. As such, parenting programs that incorporate content adapted to parent needs, such as support that is relevant and delivered when it is required, may have higher engagement.

Moderate evidence from systematic reviews, interviews, and cross-sectional studies suggests that offering some tailoring within online parenting programs, whereby content is modified to suit individual contexts, improves engagement [42,46–48]. A systematic review of parenting programs (25 studies, combined N = 5,215) concluded that tailored parenting programs had higher retention and engagement amongst underserved parents [46]. For example, fathers interviewed in focus groups suggested that they would likely find content related to a specific parenting problem preferential to generic resources [47]. Similarly, a survey of Australian parents found that programs with tailored content were among the most useful when compared to other approaches such as parenting forums or structured online programs [42]. Finan et al [48] also found that parents were more likely to engage with parenting programs when they had some control over program access (i.e., location) and usage (i.e., timing), as employed in adult mental and public health programs via ecological momentary interventions [9,49,50]. These studies indicate that tailoring content in parenting program may lead to improvements in retention and engagement.

Ecological momentary interventions offer potential for providing tailored program content in the moment. These approaches provide treatments or resources to individuals at a time and place most convenient to them [9]. Ecological momentary interventions typically collect ‘real-time’ responses in order to identify when an individual is experiencing, or is at risk of experiencing, particular symptoms, and then provide strategies to alleviate or prevent those symptoms by immediately delivering a brief intervention [51]. For example, smoking cessation interventions aim to decrease or terminate smoking habits by providing a brief in-the-moment intervention that disrupts individuals’ urge to smoke. Interventions can include strategies such as watching videos [52], meditation [52], tips [52,53], and supportive messages [52,53]. A meta-analysis of smoking cessation apps (10 studies, combined N = 2,391) concluded that ecological momentary interventions were significantly more likely to promote smoking abstinence compared to controls without an ecological momentary intervention element [54].

It may be possible to apply an ecological momentary design in the context of parenting programs, using real-time data collection to allocate tailored in-the-moment support. Indeed, recent emerging evidence suggests that optimising the timing of parenting programs depending on circumstance improves engagement levels of parents [55]. In a study of 8,074 parents, Cortes et al [55] sent text messages about literacy skills to parents during a weekday or a weekend period, with high-educated parents engaging more with messages scheduled for weekdays and lower-educated parents engaging more with messages scheduled for weekends. The results reinforce that identifying when parents are most receptive to an intervention is important in predicting program efficacy. As in adult mental and public health studies, it may be that parent engagement is highest at times of parent or child emotion dysregulation or difficulty, i.e., when parents are most likely to react punitively or less effectively [56–58]. Alternatively, it is possible that dysregulated moments may not be optimal for offering parents an immediate intervention, as parents may be distracted and focused on managing their child in the moment, either unable to engage with an app or to focus their attention on learning materials. Thus, unlike adult mental and public health programs, it is not clear what the most appropriate timing would be for delivering in-the-moment support to parents.

Ecological momentary assessment (EMA) involves collecting data in the moment multiple times per day [59], and is used to inform the delivery of ecological momentary interventions. EMA is distinct from ecological momentary intervention in that it solely involves the collection of data regarding an individual’s current behaviours or thoughts, both as the person experiences them and in the environment that they are experiencing them [59]. Encouragingly, EMAs demonstrate high levels of engagement, with compliance rates above 80% in adult mental and public health contexts [60]. They are therefore ideal to use with programs that rely on accurate, timely data, like smoking cessation programs, which ask whether an individual is craving a cigarette or vape in the moment (i.e., ecological momentary assessment) and, if so, offer an intervention to alleviate the craving (i.e., ecological momentary intervention). Although the use of EMAs in parenting contexts is scarce, two studies have shown promise in improving parent engagement. Dunton et al [61] and Li and Lansford [62] utilised EMA to assess the impact of parent stress on parenting practices, finding that parents completed almost 80% of prompts on average. Dunton et al’s [61] study recruited 199 mothers of children aged 8–12 to complete two 7-day periods of EMA data collection, with survey prompts sent up to eight times per day. Li and Lansford [62] recruited 184 parent-child dyads of children with and without ADHD and prompted participants to answer one questionnaire a day for 7 days. Although both studies stated limitations due to participant self-report, both utilised EMA via smartphones, strengthening support for its use in parenting contexts. Conway, Wladis and Hachey [63] highlighted that parents of children aged under 6 years were more time poor than parents of older children, however, both study samples recruited parents of children aged at least 6 years or older on average, which may not reflect the capacity of all parents to maintain high engagement. These nascent results support the use of EMA to collect in-the-moment responses from some parents, which in turn can be used to inform the development of an ecological momentary intervention as a parenting program.

The current study sought to inform a larger project developing a parenting app by designing a valid and reliable EMA that would be delivered to parents at optimal times of the day to maximise parent engagement. An effective and well-utilised emotion regulation parenting program, underpinned by ecological momentary design, should be (a) brief, thereby reducing burden; (b) valid, thereby accurately measuring dysregulation; and (c) timely, prompting parents when they are most in need of assistance and are most likely to engage. Thus, the current study aims to:

1. Identify items for a very brief EMA which best measures in-the-moment parent and child negative affect and emotion dysregulation;
2. Determine the optimal time of day for participant engagement with smartphone prompts;
3. Assess how often and when specific difficult parenting situations occur; and
4. Establish what times of day and which parenting situations precipitate heightened levels of parent and child negative affect and emotion dysregulation.

Materials and methods

Parents

Fig 1 shows the flow of participants through the study. Of the initial 635 baseline responses, 451 parents were excluded primarily due to issues with completeness or eligibility (i.e., not residing in Australia or not a parent of a child aged 2–4 years). A further 28 were excluded due to difficulties enrolling in SEMA3, with a final 47 excluded for not logging into the SEMA3 app and therefore not receiving any EMA surveys. After exclusions, the final sample consisted of 89 parents with full baseline data who engaged with the SEMA3 platform. Three parents included in the final sample received EMA prompts but did not complete any EMA surveys.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of Participant Retention Through the Study.

https://doi.org/10.1371/journal.pone.0327179.g001

Table 1 summarises the demographic characteristics of the sample. Eligible parents were required to reside in Australia and be a parent of a child aged 2–4 years. Parents were predominantly female, highly educated, and had a medium-high income. Just under half the children were reported as female, and almost half were aged 2 years.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Participant demographic characteristics.

https://doi.org/10.1371/journal.pone.0327179.t001

Measures

In the baseline questionnaire, parents were invited to provide demographic and contact information, as well as information about their parenting, the family emotional climate, and their child’s disposition. Although not all measures were validated for use with children aged 2–4 years, measures whose items were deemed appropriate for the age range were selected. On enrolment in SEMA3, parents were asked to complete a short survey about their and their child’s in-the-moment emotion dysregulation and negative affect, as well as if they had recently encountered any difficult parenting situations. Based on prior research investigating which situations parents most often solicited advice on parenting forums [41], parents were able to select from a list of 27 pre-populated parenting situations (e.g., ‘Child resisting going to bed/sleep’; ‘Child fighting with their sibling’) or select ‘Other’ and describe the recent difficult parenting situation (see S2 Table for the full list of parenting situations). These outcomes were selected to align with a larger project aimed at developing a parenting program to improve parent and child emotion regulation. Tables 2 and 3 detail the measures and items used in the baseline questionnaire and short surveys.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Summary of 25-minute baseline questionnaire measures.

https://doi.org/10.1371/journal.pone.0327179.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Summary of 1-minute survey measures.

https://doi.org/10.1371/journal.pone.0327179.t003

Results

Item selection

Table 4 presents an overview of results according to the five criteria (sensitivity; relevance; alignment; frequency; face validity) used to evaluate the 19 short survey items. Six items were selected for inclusion in the subsequent analyses, based on the ranking assigned to each item within each measure, with the final selection including one item measuring parent negative affect, two measuring parent emotion dysregulation, two measuring child negative affect, and one measuring child emotion dysregulation.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Evaluation of short survey items’ sensitivity, relevance, alignment, frequency, and validity.

https://doi.org/10.1371/journal.pone.0327179.t004

For parent negative affect, item 1 (Upset) was selected as most appropriate. Item 1 had the strongest alignment and highest frequency. In terms of face validity, the item was considered as having flexibility in reflecting different types of parent affect. Items 3 (Ashamed) and 5 (Afraid) were highest for sensitivity, however, were not selected as more than 30% of parents reported never experiencing those emotions during the one-week study period. Item 2 (Hostile) was not selected as the research team decided it was not an appropriate emotion to measure parent negative affect, whilst item 4 (Nervous) was discounted in favour of item 1.

For child negative affect, items 2 (Angry) and 5 (Sad) were selected as most appropriate. Items 2 and 5 had second highest and highest alignment, respectively, whilst the research team also decided that it was important to capture emotions that indicated parents and children were most at risk of emotion dysregulation. Additionally, items 1 (Depressed), 3 (Scared), and 4 (Afraid) were never reported as the child’s in-the-moment emotion by more than 45% of parents, which led to those three items being discounted.

For parent emotion dysregulation, items 1 (‘My emotions feel overwhelming’) and 2 (‘I am having difficulty controlling my behaviours’) were selected as most appropriate. Item 1 had the strongest alignment and equal-highest relevance, with strong face validity. The team considered item 1’s mention of emotions to align with the objective of an emotion regulation parenting app, and item 2 had the highest sensitivity and equal-highest relevance. Item 3 (‘I am having difficulty doing the things I need to do right now’) had equal-highest relevance and second-highest alignment, however, the research team considered that participant responses to the item may be related to factors not relevant to emotion regulation (e.g., child behaviour, other contextual factors), and it was therefore discounted. Items 4 (‘I am paying attention to how I feel’) and 5 (‘I have no idea how I am feeling’) scored poorly across almost all criteria and were therefore not considered for selection.

For child emotion dysregulation, item 3 (‘My child has difficulty controlling their behaviours’) was selected as most appropriate. Item 3 had the equal-highest frequency, and second highest and equal-second highest sensitivity and alignment. In terms of face validity, the research team considered item 3 might detect child emotional distress which parents perceived solely as behavioural problems in the moment. Item 4 (‘My child has difficulty doing things they need to do now’) had the highest relevance and alignment, however, the research team decided that the item was too broad and may not be relevant for children aged 2–4 years, and as such it was discounted. Item 1 (‘My child seems to be overwhelmed by their emotions’) had the second highest relevance and equal-second-highest alignment, and item 2 (‘My child seems to be feeling out of control’) had the highest sensitivity, however, parents less often reported the statements as being reflective of their child in the moment than item 3 (i.e., frequency), leading to both items being less preferred than item 3.

Survey engagement

Smartphone prompt response rate.

Table 5 shows when parents were most likely to respond to smartphone prompts. Parents were less likely to respond to smartphone prompts during the early morning, the start of the workday, and at night, with no major discrepancies in response rates at other times of the day.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Frequency of participant responses to smartphone prompts during the day.

https://doi.org/10.1371/journal.pone.0327179.t005

Frequency of parent and child together.

Table 6 shows whether parents reported that they were with their children while responding to a smartphone prompt. Children were most often with their parents during the morning, specifically 7:30am, and early evening (6:00 pm). In contrast, parents were least likely to be with their children at night, as well as during working hours.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Frequency of the times of day parents are with their child.

https://doi.org/10.1371/journal.pone.0327179.t006

Timing of difficult parenting situations.

Table 7 reports the 20 parenting situations (of 27 tested) rated as the most common by parents, and the times at which they were most often reported. Ratings for some situations were influenced by the time of day. For example, a child resisting going to bed was reported as a difficult parenting situation almost 70% of the time from 7:30 pm onwards, whereas a child whining for more screen time did not fluctuate significantly throughout the day.

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Most common parenting situations and their frequency during the day.

https://doi.org/10.1371/journal.pone.0327179.t007

Heightened parent and child negative affect and emotion dysregulation.

Table 8 reports mean scores for parent and child negative affect and emotion dysregulation associated with the 20 most common difficult parenting situations. Parents reported the highest levels of parent negative affect when their child refused to hold their hand whilst crossing the road and highest emotion dysregulation when their child resisted going to bed. Meanwhile, parents reported that their child showed highest levels of negative affect when the child did not want to leave the parent’s side and highest emotion dysregulation both when their child did not want to leave the parent’s side and when the child refused to put their shoes on.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. Most common parenting situations with means and 95% confidence intervals of parent and child negative affect and emotion dysregulation.

https://doi.org/10.1371/journal.pone.0327179.t008

Table 9 shows the mean scores of parent and child negative affect and emotion dysregulation throughout the day. Parents reported heightened levels of negative affect and emotion dysregulation in the evening, whereas child negative affect was higher in the late afternoon and emotion dysregulation was higher in the evening.

Download:

• PNG
  larger image
• TIFF
  original image

Table 9. Comparison of times of day with means and 95% confidence intervals of parent and child negative affect and emotion dysregulation.

https://doi.org/10.1371/journal.pone.0327179.t009

Discussion

Findings from the current study will inform the development of an in-the-moment parenting smartphone app to improve emotion regulation in children aged 2–4 years. The study used an EMA approach to collect information from 89 parents with a child aged 2–4 years to determine the (1) optimal item selection for measuring in-the-moment parent and child negative affect and emotion dysregulation, (2) timing of prompts, (3) selection of daily experiences of difficult parenting situations in an Australian setting, as well as to (4) establish times of day and parenting situations that precipitate heightened levels of parent and child negative affect and emotion dysregulation.

For aim 1, six items of 19 tested were selected to measure parent and child negative affect and emotion dysregulation: ‘Upset’ (parent negative affect); ‘Angry’ and ‘Sad’ (child negative affect); ‘My emotions feel overwhelming’ and ‘I am having difficulty controlling my behaviours’ (parent emotion dysregulation); and ‘My child has difficult controlling their behaviours’ (child emotion dysregulation). No specific recommendations currently exist to guide the development of an EMA measure, but previous research using brief EMA measures utilises a wide range of approaches to assess reliability and validity [83,84]. Thus, our study extends the EMA literature by providing a coherent and detailed assessment framework including five elements: sensitivity, relevance, alignment, frequency, and face validity. While consideration was given to the psychometric strengths of each item (sensitivity, relevance, alignment, and frequency), the face validity of each item was given the greatest weight due to the lack of existing parenting EMA studies. This approach of prioritising face validity in item selection for EMA measures has been utilised by other studies that sought to develop very brief measures [83,85].

Aim 2 of the current study sought to identify when parents would be most likely to respond to smartphone prompts and to be with their children. Parents were least likely to respond to prompts at 6:30am, 9:00am, and 9:00 pm. While parents were not asked why they were unable to complete prompts, it may be that families were not awake or busy with children at 6:30am. Parents may be beginning their workday at 9:00am, and at 9.00 pm might be tired from the day or avoiding technology. In general, parents reported most often being with their child at 7:30am and 6:00 pm, however, this varied greatly between weekdays and weekends – on weekends, parents reported being with their child 10−30% more often during the hours of 9:00am-4:30 pm. As more than 70% of the sample was employed, and 80% of the children in the sample were enrolled in early childhood programs, parents were likely to be at work during the hours of 9:00am-4:30 pm on a weekday, rather than with their child. Parents may have indicated that they were not with their child at 6:30am and 9:00 pm because their child was asleep, even if they were in the same house. Results showed the average participant response rate was 81%, similar to other studies using EMA in parenting contexts, [61,62], and EMA more generally [34]. Although the result should be interpreted with caution as participants were only involved for one week and were paid to participate, this response rate is surprising considering previous evidence that suggests parents with children under 6 years are more time poor compared to other parent groups [63]. Overall, results support the use of EMA to improve parent engagement and suggest the need for tailoring the timing of EMA prompts to maximise the opportunity for parents to respond.

Aim 3 investigated the most common difficult parenting situations and times of day when they occurred. The five most common situations included difficulties with food, bedtime, sibling conflict, transitions, and screens, however, the time of day in which they occurred varied by situation. Situations related to routines or daily rhythms were more often reported in the morning or evening, while other situations were reported evenly throughout the day. For instance, children fighting with their sibling was reported relatively evenly across the course of the day, while children resisting going to bed was typically reported in the two latest time periods (7:30 pm and 9:00 pm). Parenting programs often use a one-size-fits-all approach to content, without consideration for the unique circumstances of families [42,47]. Such an approach results in parents disengaging from programs, as they do not feel that the resources provided reflect their individual needs as parents and families.

Aim 4 of the current study sought to identify the types of parenting situations and times of day that led to heightened parent and child negative affect and emotion dysregulation. In general, there was no pattern of dysregulation for parents or children when it came to parenting situations. Some situations that were reported infrequently resulted in high levels of dysregulation, while other more frequent situations did not elicit a similar response. For example, parents responded that it was rare for children to resist holding their hand when they crossed the road, however, on the occasions that it did happen, parents showed higher levels of negative affect compared to other situations. Conversely, when parents reported that their child was fighting with their siblings, they did not report higher dysregulation in either themselves or their child compared to other situations. It may be that parents and children naturally learn coping strategies for situations which are more common, thereby reducing in-the-moment dysregulation. Further to this, parents and children did not necessarily show heightened regulation in the same situations. Children appeared to exhibit highest levels of negative affect and emotion dysregulation in situations where they needed to leave their parents, whereas parents reported highest negative affect when their child did not hold their hand whilst crossing the road. These discrepancies could be due to parents’ and children’s different perceptions of danger, with parents more fearful of the immediate danger or fast-moving traffic and children more uneasy about being away from their parents.

Compared to parenting situations, time of day appeared to show a more consistent pattern whereby both parents and children showed higher dysregulation from the late afternoon onwards and seemingly peaking at around 7:30 pm. This suggests that coping for both parents and children is lower later in the day. This may be because all parties have fewer emotional resources after work or childcare, which may then be exacerbated by evening routines that cause conflict. This is emphasised by the fact that children resisting going to bed was the second-most reported parenting situation, which also resulted in high levels of negative affect and emotion dysregulation for all. The final reported time of day, at 9:00 pm, showed a noticeable decrease in dysregulation, potentially suggesting that once children were asleep, parents were able to report feeling less dysregulated. The results reinforce that the development of parenting programs would benefit from consideration of the contexts in which parents request an intervention and the times at which they are most likely to require support.

Adult mental and public health programs have had success in offering ecological momentary interventions that provide contextual, in-the-moment resources. Our findings offer concrete information on how similar designs could be implemented within a mental health prevention context, such as offering parenting support via ecological momentary interventions to young families. Our study was specifically designed to inform the development of a parenting program that aims to improve parent and child emotion regulation using an ecological momentary design, although the findings are likely to have relevance to any program using a similar design. EMAs provide in-the-moment data that reflect an individual’s experience of their environment in real time, allowing for parents to provide more accurate responses about their and their child’s affect and emotions. Evidence suggests that parents prefer programs with content tailored to their context [46], and the diversity in the types of difficult situations that parents reported and the times during which they occurred throughout the current study indicates that tailoring content to reflect all kinds of family situations is warranted when developing future parenting programs.

In considering whether to use EMA for parenting programs, the current study examined whether a smartphone-based approach would result in higher levels of engagement by parents and how programs could maximise the efficacy of their content. Our findings offer concrete information on how similar designs could be implemented within a mental health prevention context. In understanding the barriers that families face with in-person programs, online approaches can help to mitigate some of these obstacles whilst reaching families that are most substantially limited by them. In turn, with parents indicating their desire for content that more accurately reflects their own personal circumstances, providing opportunities for support and learning both when parents can access them and in times when their needed and convenient may precipitate stronger engagement and improved outcomes. Parents expressed a desire for content that accurately represents their circumstances, with flexible access to support and learning opportunities. Our findings align with evidence from Finan et al [48], Hansen, Broomfield and Yap [46], and Sicouri et al [47], indicating that parents prefer programs tailored to their context. The diversity in reported parenting challenges – both in type and timing – suggests that future programs should reflect a wide range of family situations. Ensuring availability when needed and at convenient times may enhance engagement and improve outcomes.

Conclusion

Parenting programs are an effective means of modifying parents’ skills and behaviours and improving children’s outcomes, but have struggled with low population reach and engagement. Online programs have offered potential to extend the reach and engagement of parenting programs, but so far have also struggled with low engagement and high attrition. The proliferation of smartphones amongst adults in Australia represents an opportunity for parenting interventions to provide timely and contextual resources to parents, but data is needed to guide intervention design. Findings from the current study provide concrete information about potential for tailoring parenting online and ecological momentary interventions to parents with young children. This study used a pre-existing set of short measures to develop a 1-minute ecological momentary assessment survey for parent and child negative affect and emotion dysregulation to then identify when parents were most likely to engage with smartphone prompts, when they were most often with their children, and the frequency at which different difficult parenting situations arose.

Supporting information

Acknowledgments

The authors would like to acknowledge all the participants who offered their time and efforts to this project. In addition, Tomer S. Berkowitz would like to acknowledge Lauren and Liam for their help with code, Chris and Liam for providing specific code for some of the analysis, and Bonnie for all her work getting the project up and running.