Does Engaging in a Group-Based Intervention Increase Parental Self-efficacy in Parents of Preschool Children? A Systematic Review of the Current Literature

The preschool years, defined in this review as 0–6 years of age in line with international school starting ages (Sargent et al. 2013), are a period of rapid physical and psychological development for children. The need to implement early intervention strategies to give children the best opportunity to strive and pre-empt later personal and societal level problems has been widely recognised (e.g. Allen 2011). To this end, the UK Government emphasised the need for evidence-based preventative interventions aimed at increasing the parental competence of parents of preschool children (Allen 2011; Leadsom et al. 2014).

Parents are integral in shaping their child’s physical, emotional and social environment and thus their development. According to Waldfogel and Washbrook (2011), parental behaviours play a significant role in a child’s psychosocial development even after controlling for demographic characteristics. Lack of positive attention from parents paired with inconsistent and inappropriate discipline has been found to be predictive of anti-social behaviour, conduct disorder and criminality in later life (Farrington 2005). Conversely, positive reinforcement, responsiveness, warmth and positive affect have been associated with positive child developmental outcomes (Gardner et al. 2003; Shaw et al. 1998; Zhou et al. 2008). As parenting style has been shown to be adaptable to change (Taylor and Biglan 1998; Wykes et al. 2008), a pivotal mechanism for early intervention is through parents and promoting effective parenting.

According to Bandura (1982, 1997), self-efficacy has been defined as an individual’s beliefs about their capabilities to produce desired levels of performance to influence events that affect their lives (Bandura 1982). Thus, self-efficacy is central to conducting behaviour and it influences behaviour change. Parental self-efficacy is a subcategory of general self-efficacy and has been broadly defined as the expectation a parent holds about their ability to parent successfully (Jones and Prinz 2005). Strong evidence for a link between parental self-efficacy and parental competence has been found, with some evidence that higher levels of parental self-efficacy are related to more effective parenting and better child outcomes (see Jones and Prinz 2005 for a review). In addition, high maternal self-efficacy has been linked to increased sensitivity and responsiveness towards their child in longitudinal studies (Dumka et al. 2010; Teti and Gelfand 1991) as well as to increased maternal warmth in cross-sectional studies (Izzo and Weiss 2000). These factors in turn have been found to predict decreases in child aggression (Jones et al. 2008) and positive social-emotional development in children in longitudinal studies (Page et al. 2010). In contrast, a link has been found between lower parental self-efficacy and higher dysfunctional parenting, including laxness and over-reactivity, in several cross-sectional studies (e.g., Gross et al. 1999; Sanders and Woolley 2005).

As self-efficacy is not a fixed personality trait but a dynamic process modified by task and situational demands as well as changing individual factors (Bandura 1997), parental self-efficacy should be a crucial consideration when assessing parenting interventions aimed at increasing parental competence. Support has been found for multiple mechanisms through which parental self-efficacy influences parental behaviour, indicating parental self-efficacy can be an antecedent, consequence and mediator of parenting.

High parental self-efficacy has been linked to positive parenting strategies and behaviours (Coleman and Karraker 1998). When parents felt competent in their ability to parent, they were likely to use more effective parenting practices, which foster positive developmental outcomes (Bloomfield et al. 2005). According to Coleman and Karraker (1998), self-efficacy beliefs influence parenting behaviours via a dynamic interaction of affective, motivational, cognitive and behavioural pathways. Low self-efficacy can also have a direct impact on behaviour through inhibiting the acquisition of new skills and suppressing existing skills (Bandura 1982), which is particularly pertinent when considering how to optimise the acquisition and use of positive parenting skills. Effective parenting leads to enhanced feelings of efficacy in a parent (Bandura 1997). A number of factors influence parental self-efficacy, including social support, infant temperament and maternal mental health (Cutrona and Troutman 1986; Leahy-Warren and McCarthy 2011; Leerkes and Burney 2007). Consequently, it is possible to identify particular groups of parents that are at risk of experiencing low parental self-efficacy. Parental self-efficacy may also mediate the effects of depression, perception of infant temperament and social support on parenting competence in parents of young children (Teti and Gelfand 1991). It has also been found to act as a buffer against the impact of adversity (e.g., Ardelt and Eccles 2001), which suggests that parental self-efficacy is a potential area of intervening and attenuating the effects of non-manipulatable variables, such as temperament. The evidence presented so far highlights a complicated relationship between parental self-efficacy and parenting behaviours, but nonetheless points to parental self-efficacy as an important factor to be targeted in parenting interventions.

Parental self-efficacy has been measured exclusively via self-report questionnaires and at different levels: general, task specific and narrow domain self-efficacy (Jones and Prinz 2005). General parental self-efficacy measures ask a parent to comment on how competent they feel in the parenting role without focussing on specific tasks. Task-specific measures calculate global parental self-efficacy by focusing on the perception of competence over a range of discrete parenting tasks; for example, discipline and soothing a baby. Narrow-domain measures concentrate on perceptions of competency in one parenting domain; for example, involvement in school-related activities. In a study assessing the relationship between maternal self-efficacy, dysfunctional discipline style and child conduct problems, Sanders and Woolley (2005) found differing results depending on the type of measure used and the domain assessed. Parenting practise was best predicted by scores on task-specific measures of parental self-efficacy.

Given the mediating role of self-efficacy between knowledge and behaviour (e.g., Bandura 1982, 1997), the impact of group-based programmes designed to impact on parenting skill on parental self-efficacy is an area of increased interest. Perceptions of self-efficacy also determine the amount of effort an individual expends and how long they persevere in the face of adversity (Bandura and Schunk 1981). Low self-efficacy can inhibit the acquisition of new skills and suppress existing skills (Bandura 1982). This is particularly relevant to parenting interventions: increasing skill and knowledge may only lead to behavioural change if a parent also has sufficient confidence in their abilities. Indeed, Grusec et al. (1994) noticed that parents with low parental self-efficacy were not able to put parenting knowledge into practice. Coleman and Karraker (1998) suggested that traditional parenting interventions focussing on knowledge and skills alone might not suffice. Thus, in order to optimise parenting quality it may be necessary to ensure parenting groups also increase parental self-efficacy. According to Sanders and Mazzucchelli (2013), parental self-efficacy is an important element of a broader capacity of self-regulation, important for nurturing positive parenting practises. They argued that integrating a focus on parental self-regulation into parenting interventions would allow parents to become self-sufficient in creating and maintaining change (see Sanders and Mazzucchelli 2013, for a full discussion).

Much of the research into the effectiveness of group-based interventions for parents has focussed on parenting programmes (see Barlow et al. 2012; Nowak and Heinrichs 2008, for reviews). Research has been conducted into groups based on a range of theoretical perspectives, including behavioural, cognitive-behavioural, psychodynamic and social learning theory. Despite differing in orientation, parenting programmes have been found to exhibit many commonalities in their implementation (Kazdin 1997). Often, although not always, sessions involve some elements of video vignettes, role-play or an opportunity to practise new techniques, interaction coaching, didactic teaching and group discussion to help parents develop the parent–child relationship elicit their own problem-solving skills, and provide an environment within which to practise these new skills.

Several previous reviews have found evidence that group-based parenting programmes were an effective intervention for reducing child problem behaviour (Dretzke et al. 2005; Taylor and Biglan 1998; Wyatt Kaminski et al. 2008), improving positive parenting, reducing harsh parenting practises, improving emotional and behaviour adjustment in children (Barlow et al. 2005, 2011) and improving parental psychosocial adjustment (Barlow et al. 2012). Additional benefits of group-based interventions include increased access to services and cost effectiveness, especially for children with challenging behaviour (Edwards et al. 2007; Scott et al. 2001). Five influential factors that construct and enhance self-efficacy have been identified: previous experience, vicarious experience, mastery experience, verbal (social) persuasion and psychological state (Bandura 1977, 1997). Group-based programmes are the most common form of parenting interventions and have the potential to offer a supportive environment to utilise these factors to influence parental self-efficacy. For example, verbal persuasion is more likely to occur in a social setting such as a group-based intervention. Most parenting-interventions are run in a group-based setting.

Parenting groups are not the only group-based programmes available to pre-school parents. Many structured group activities are available in the UK, such as music groups, swimming groups and sensory groups. These share many similar characteristics to parenting groups albeit the content is different. Similarities include being facilitated by an instructor, incorporating a taught element for parents and a focus on developing parent-infant relationships and parenting skills. Often participation in these group-based activities provides the parent and child with an instant opportunity to practise new skills, alongside didactic teaching, parental discussion and problem solving. Key concepts identified by Kane et al. (2007) as important in providing helpful and meaningful parenting programmes (e.g., the acquisition of knowledge, skills and understanding alongside social support) are also present in non-parenting programmes.

A Cochrane review examined the impact of group interventions on parental psychological functioning including depression, anxiety, general self-esteem and parental self-efficacy (Barlow et al. 2012). Overall, the review concluded that group-based parenting programmes significantly increased short-term psychological wellbeing in parents, including significant increases in parental self-efficacy. However, although parental self-efficacy was examined as one factor in this review, the impact of group-based interventions on parental self-efficacy specifically in parents of preschool children was not the focus of their review. Thus, the aim of this paper is to critically examine the literature on the impact of group-based interventions for parents of preschool children on parental self-efficacy and to assess the methodological qualities of each identified study. As the preschool years have been identified as a key period for intervention, we focussed on studies of parents of children up to 6-years old.

The quality of the evidence, the impact of the type of parental self-efficacy measure used and the role of potential moderating and mediating factors that influence whether group-based interventions impact on parental self-efficacy were also examined.

Overall the majority of the studies reviewed reported a significant positive impact of group-based interventions on parental self-efficacy in parents of children less than 6 years of age, regardless of whether a task-specific or general measure of parental self-efficacy was employed. However, the effect sizes for the interventions did appear to differ; the majority of studies which utilised task-specific measures of parental self-efficacy found medium to large post-intervention effect sizes (0.42–1.25), while small to medium effect sizes (0.26–0.74) were found in studies using general measures of parental self-efficacy.

Follow up data were reported between 6 weeks and 2 years post-intervention. Several studies only reported follow up data for the intervention group, thereby limiting the conclusions that can be drawn. Six studies included follow up data at least 6 months post-intervention for both intervention and control groups. Three studies found significant differences in parental self-efficacy at follow-up (Gross et al. 2003; Miller-Heyl et al. 1998; Tucker et al. 1998), whereas no difference was noted in the other three studies (Breitenstein et al. 2012; Cunningham et al. 1995; Gross et al. 2009). In the study by Breitenstein et al. (2012), while improvements in parental self-efficacy were maintained in the intervention group, the control group also improved over time. Furthermore, detailed, long term research is required to explore the factors underlining the long term impact of group-based interventions on parental self-efficacy.

Increases in parental self-efficacy were seen in interventions ranging from one to 15 sessions. The finding that single session interventions can significantly increase parental self-efficacy has substantial potential clinical importance in terms of the ability to disseminate time and cost effective interventions to many parents. The four studies looking at single session interventions focussed on specific problem situations, such as mealtimes and shopping. All recruited volunteer samples of parents of children without a diagnosed behavioural difficulty so greater investigation with a range of difficulties and populations would be beneficial. Group-based interventions, regardless of the length, incorporate, and draw upon, the five factors Bandura (1977, 1997) identified as important for enhancing self-efficacy; previous experience, vicarious experience, verbal persuasion, and have been found to enhance physical and psychological wellbeing (Leahy-Warren 2005; McVeigh 1998). Additional, direct comparison of brief versus longer interventions would enable an exploration of the important elements of both interventions and whether these are the same in brief and longer interventions. The long-term sustainability of the impact of single session interventions requires further investigation as none of the reviewed studies reported control group comparison data at follow up. However, the maintenance of increased parental self-efficacy in the intervention groups after one session at 6 months is encouraging. This short, time-limited approach may fit well with the tiered model of early interventions recommended by the UK Government of universal support offered to all parents before more intensive support is focussed on “at risk” families (Allen 2011; Leadsom et al. 2014).

Few studies considered if the effectiveness of group-based interventions to improve parental self-efficacy may be moderated by individual characteristics. Ethnicity was found to be a moderating factor in one study (Breitenstein et al. 2012), but not in another study (Miller-Heyl et al. 1998). Previous research has found substantial similarity in the relationship between parental self-efficacy and parental competence across ethnic groups (see Jones and Prinz 2005).

Fathers were a significantly under-reported group in the reviewed studies, with only one study reporting paternal self-efficacy (Gross et al. 1995). Yet the available research consistently highlights fathers report lower parental self-efficacy than mothers, suggesting they are an important group to target (Elek et al. 2003; Froman and Owen 1989; Hudson et al. 2001; Reece and Harkless 1998). The lack of a positive significant impact of group intervention on paternal self-efficacy compared to maternal self-efficacy found by Gross et al. (1995) implies different approaches may be necessary to increase paternal self-efficacy. Indeed, group interventions have been found to be effective in increasing paternal self-efficacy when delivered in an online forum or father-only group format (Hudson et al. 2003; McBride 1990, 1991).

Few studies directly examined the relationship between parental self-efficacy and parent and child behaviours. Thus, conclusions about whether improving parental self-efficacy is a mechanism by which group-based interventions facilitate changes in parenting competencies cannot be drawn. As such further research explicitly investigating the role of group-based parenting interventions in increasing parental knowledge and self-efficacy and how this relates to behaviour change is warranted to further build on the theory by Bandura (1977, 1982, 1997). Some support, however, was found for a correlation between increased parental self-efficacy and an increase in some positive parenting skills (Gross et al. 1995; Miller-Heyl et al. 1998). Only two studies assessed the relationship between changes in parental self-efficacy and reported child behaviour (Gross et al. 1995, Pisterman et al. 1992). While Gross et al. (1995) reported parental self-efficacy correlated with reductions in child behavioural problems, Pisterman et al. (1992) found parental self-efficacy was unrelated to changes in child behaviour in children. One possible explanation for the difference in findings may be that the children in the study by Pisterman and colleagues had greater behavioural problems as they had a diagnosis of ADDH. Further research into the mechanisms of change is needed. Does parental self-efficacy impact on child behaviour because of the impact on particular parenting behaviours, or an increase in consistency or a greater perceived ability to manage problem behaviours? How the intensity of behavioural problems mediates the relationship between parental self-efficacy and child behaviour also requires investigation. Independent assessment of child behaviours alongside parental self-report is needed to explore this.

Overall, evidence was found for a significant benefit of group interventions on parental self-efficacy. However, the methodological quality of the reported studies is important to consider. Eight out of 15 studies were rated as having inadequate quality on the CTAM. An inspection of the seven studies rated the highest quality showed that, of the six which reported post-intervention findings, four found a medium or large effect of intervention (Adamson et al. 2013; Gross et al. 2003; Hayes et al. 2008; Morawska et al. 2014) and two found no effect (Breitenstein et al. 2012; Cunningham et al. 1995). Furthermore, the effects were lost in two studies when more conservative intention-to-treat analyses were conducted (Adamson et al. 2013; Hayes et al. 2008). One possible explanation for the differing results could be differences between the sensitivity of the measures used to identify changes in parental self-efficacy as four different measures were used across the six studies. Of note, the two studies which used general measures of parental self-efficacy either found no effect or the effect was lost when an intention-to-treat analysis was conducted (Cunningham et al. 1995; Hayes, et al. 2008). As reported earlier, effect sizes were generally greater in studies utilising task specific measures of parental self-efficacy and it may be these measures are more sensitive. No clear pattern was seen in terms of sample characteristics, the content of the group or length of intervention. However, the level of description of the content of the intervention varied between papers, therefore such differences may not have been identified. Further research is clearly indicated.

Several limitations of reviewed literature were identified. Parental self-efficacy was not a main outcome measure in the majority of the studies reviewed which as noted, influences the level of information provided and the conclusions that can be drawn. Few studies also explored potential moderators. As such, further empirical testing was not possible in this review; which is noted as a limitation. Secondly, the methodology of eight out of 15 studies reviewed was judged to be inadequate. Studies with poor methodological quality often inflate the effects of psychological interventions (Wykes et al. 2008). Thus, the conclusions of the studies judged to be inadequate on the CTAM must be interpreted with caution. Thirdly, only a minority of the studies reviewed (seven studies) compared both intervention and control groups beyond the immediate post group evaluation. Additionally, the follow period was limited to 6 months post intervention in all but four studies. Greater longitudinal evaluation is required to establish any lasting impact on parental self-efficacy. Fourthly, all studies utilised self-report questionnaires as the only measure of parental self-efficacy. This is a limitation of the research area as a whole. There is not an independent measure of parental self-efficacy that can be completed by an independent rater, but this likely reflects the highly personal nature of the concept. Finally, variation exists in how the construct of parental self-efficacy is operationalised, both in terms of general versus task-specific measures and between individual questionnaires, which makes synthesis across studies more difficult. However, in a previous literature review, Jones and Prinz (2005) concluded that despite the variation in conceptualising parental self-efficacy, the different measures broadly tapped into the same concept. Nonetheless they cautioned that systematic trends in findings may be caused by different measurement strategies and this concern must be taken into account with the literature reviewed here, especially in terms of the differing effect sizes reported for general and task-specific measures. Greater consensus on the concept of parental self-efficacy and standardisation of measurement tools would aid future studies and develop this area of research.

A meta-analysis was not completed for this literature review for several reasons. While meta–analysis can be less prone to bias than narrative literature reviews (Teagarden 1989), this is only true when procedural differences between studies are low and methodological quality is consistently high. Using meta-analysis when these assumptions are violated can lead to poor, and at worst, harmful conclusions (Bailar 1995). Due to the differences between interventions (e.g. length of intervention, age range of children) and the varying quality of studies reviewed, meta-analysis was deemed inappropriate. Nevertheless, effect sizes, where possible, were calculated to aid comparison amongst studies.

The paucity of research into paternal self-efficacy highlights this is an area which requires future research. One possible area of investigation is whether different approaches are required to increase parental self-efficacy in mothers and fathers via group-based interventions. In addition, the finding that brief, single session programmes may increase parental self-efficacy warrants further investigation, including the association of this delivery format and its long-term benefits. Research so far into this format has focused on non-clinical, volunteer samples and whether similar results for single session interventions would be replicated in clinical populations remains to be seen. This review was not restricted to parenting interventions; however, no studies evaluating the impact of non-parenting group interventions on parental self-efficacy were identified in the literature search, yet there are many non-parenting group activities aimed at parents such as baby and toddler swimming, baby massage and baby and toddler music groups. These groups share many of the characteristics of parenting interventions without the explicit focus on specific parenting skills and are perhaps less stigmatising than attending a specific parenting class (Johnson et al. 2005). Their potential to improve parental self-efficacy is an area that warrants investigation. Future research also needs to examine if the mechanisms underpinning parental self-efficacy differs across child age from birth to pre-school age.