3.2. Main Analysis
The descriptive statistics of the measures across the four time points of assessment (Time 1-4) and for each group separately (PC and CP) can be found in
Table 1. Hence, referring to Parenting styles for the PC group, we can see that Authoritarian and Permissive styles keep decreasing, whereas the Strict style seems to be constant across time. Moreover, although the Democratic style seems to increase slightly from Time 1 to 2, afterward, it remains constant. This pattern is different in the CP group, wherein the first two time points, the Democratic style remains constant, but there is an increase from Time 2 to 3. As for the Authoritarian and Permissive styles, they both decrease from Time 2 to 3. These results reveal that in general, all parents use more democratic and less authoritarian and/or permissive parenting practices after their participation in the parent training, since the change is obvious from Time 1 to 2 for PC group, and from Time 2 to 3 for CP group.
As far as Parenting Stress is considered, although a decreasing pattern across time seems to exist for all subscales, this pattern is more evident within the PC group than the CP group. The parental self-efficacy, as measured by the PSOC scale, is increasing for the CP group as time passes, whereas there is an increase from Time 1 to 2 for PC group, but it remains constant from Time 2 to 3.
Regarding the results from the ADHD-RS, the Inattention score, for the PC group, seems to keep decreasing from Time 1 to Time 3, but it remains constant from Time 3 to 4, whereas Impulsivity/Hyperactivity score reveals a decreasing pattern in all time points. As for the CP group, a more consistent decreasing pattern exists for both subscales. In general, all parents report fewer inattention and impulsivity/hyperactivity symptoms in their children after parents’ and/or children’s participation in the intervention program.
Figure 2 includes the means of parenting stress (total score), parental self-efficacy (parenting sense of competence; total score), and the two subscales of ADHD (Inattention and Impulsivity/Hyperactivity), across time and groups. Obviously, as hypothesized, all the variables are decreasing as time passes, except for parenting sense of competence, which is increasing. Note also that there is no evidence of interaction between time and group. Therefore, our hypothesis that parents of PC group would report lower levels of parenting stress and higher levels of parental self-efficacy at Time 2 assessment, compared to CP parents, was not supported.
The correlations between the variables of our study, separately for each time point and group (PC and CP; the assessment of significance must be treated with caution due to the small sample sizes), were calculated (
Supplementary Table S1). The positive correlation between Democratic parenting style and Parental Distress at Time 1 (
r = .341) for the PC group, becomes negative as time passes, while the negative correlation observed between the two variables, for the CP group, is further decreasing. Also, the negative correlation between the Democratic parenting style and the Parent-Child Dysfunctional Interaction score at the beginning of the program (
r = -.292,
r = -.216 for Time 1 and 2, respectively) is becoming much smaller after Time 2 for PC group; for the CP group, there is also an overall decrease from Time 1 to 4. The almost zero correlation between the Democratic parenting style and the Difficult Child subscale, becomes negative as time passes, for the PC group, but it seems to remain constant over time for the CP group. The same pattern seems to occur between the Democratic style and the total score of the parenting stress scale, for the PC group. However, this is not the case for the CP group, where the negative correlation between the two variables is further decreasing as time passes. Moreover, the negative correlation between Democratic style and Impulsivity/Hyperactivity at Time 1 (
r = -.743) for the CP group becomes almost zero at Time 4 (
r = .079). On the other hand, there is a positive correlation between the Democratic style and the total score of PSOC for both groups in each time point.
As far as Authoritarian style and its correlations with the subscales of PSI are considered, there are positive correlations between the Authoritarian style and the Parental Distress and Parent-Child Dysfunctional Interaction subscales for both groups (PC and CP). The negligible correlation between Authoritarian style and Difficult child subscale, becomes negative as time passes, for CP group. Reversely, the positive correlation between Authoritarian style and Impulsivity/Hyperactivity at Time 1, for CP group, becomes negligible at Time 4. The same pattern occurs between the Authoritarian style and Inattention score for PC group. Also, for both groups, there is a positive correlation between Authoritarian style and the total score of parenting stress, while the correlation with parental self-efficacy is negative.
The Permissive style seems to be more positively correlated with Parent-Child Dysfunctional Interaction, Difficult Child and Total score of PSI, as time passes, for both groups. In contrast, its correlation with total score of PSOC scale and Inattention, is becoming negative, especially for PC group.
The relatively small correlation of Inattention with the variables of PSI (Parental distress, Parent-child dysfunctional interaction, Difficult child and Total score), at first and last time point, are combined with the moderate positive correlation at Time 2 and 3, for PC group. Also, consistent positive correlations exist, between Impulsivity/Hyperactivity and the subscales of PSI, over time, for PC group. For CP group, however, the positive correlation at Time 1, between Inattention and Parental Distress, becomes negative at the end of the study (Time 4), while the negative correlation between Inattention and Parent-Child Dysfunctional Interaction at Time 1 and 2 becomes positive at Time 3 and 4. Moreover, the positive correlation between Inattention and Difficult Child at the start of the study became even larger at Time 4 (from r = .406 to r = .841). As far as Impulsivity/Hyperactivity, for CP group, is considered, it is worth mentioning the change of direction of the correlations with Parental distress and Parent-Child Dysfunctional Interaction, from positive at Time 1 (r = .255 and r = .238, respectively) to negative at Time 4 (r = -.323 and r = -.121, respectively). The same pattern is observed between Impulsivity/Hyperactivity and Total Score of PSI, for the CP group.
As far as Parenting Sense of Competence is considered, for PC group, the negligible negative and positive correlation with Inattention at Time 1 and 4 respectively, are combined with moderate negative correlations at Time 3 and 4. Conversely, the negative correlation with Impulsivity/Hyperactivity at Time 1 increases by Time 4 (from r = -.350 to r = -114). For the CP group, however, the negative correlation at Time 1 (r = -.266) becomes positive by Time 4 (r = .176).
The most important correlations, according to our aims, to be mentioned is the overall positive correlation of Democratic style with the parental self-efficacy (PSOC Total score) over time, and the mainly negative correlations with the variables of PSI for the parents of both groups. In contrast, there are positive correlations between the Authoritarian and Permissive parenting styles with parenting stress, and negative correlations with parental self-efficacy, for both groups, over time.
Moreover, as one would expect, there are mainly positive correlations between the subscales of PSI across time, for both groups. Conversely, there are mainly negative correlations between the parental self-efficacy (PSOC) with Parental Distress, Parent-Child Dysfunctional Interaction and Total Score of PSI. As far as ADHD-RS is considered, there are mainly positive correlations between the two subscales (Inattention and Impulsivity/Hyperactivity) with Difficult child and Total Score of PSI, while there are mainly negative correlations between the two subscales and PSOC for both groups across time.
The next step would be the assessment of the effect of the intervention (time), group (“PC” and “CP”), training participation (“yes” and “no”), age, and parent’s gender, on our dependent variables (parenting stress, parental self-efficacy, and core symptoms of children’s ADHD). Note that, although the correlation coefficients or the difference of means across groups or time, provide us with useful perspectives in assessing the relationship between variables, they both ignore the multivariate nature of our data. Hence, we initially fit a set of different marginal models (due to the potentially correlated errors), according to the assumed structure of the covariance matrix of errors, and then, the best model is given by the AIC and BIC criteria [
86,
87,
88]. Two interaction terms were among the exploratory variables: between time and group, and between time and training participation.
Table 2 contains the estimated marginal models; note that, following a backward selection procedure, all the interaction terms have been excluded since their contribution is not significant. Thus, for each model, we have resulted in keeping only time and group, as exploratory variables. There is no evidence of a severe violation of model assumptions (i.e., linearity, normality of residuals, and independence between residuals and exploratory variables). All the assumptions have been checked based on the properties of the normalized residuals. Given the regression coefficients of Time, we can see that as time passes, the estimation for Parenting stress, Inattention, and Impulsivity/Hyperactivity decrease (keeping the group fixed). Therefore, we expect a decrease of 5.31 units in stress, on average, from one time point to another. Also, we expect, on average, a decrease of 1.11 units in Inattention and 1.47 units in Impulsivity/Hyperactivity, from one time point to another. In contrast, the effect of Time on Parental self-efficacy is positive and equal to 3.10 units on average.
Furthermore, the effect of the PC group on parental self-efficacy, as measured by the total score of the PSOC scale, and Impulsivity/Hyperactivity is significant. Thus, the estimation for the PC group for Parental self-efficacy is lower than that of the CP group, by 5.94 units, on average, while the estimation for Impulsivity/Hyperactivity is higher by 6.60 units, on average (and keeping time fixed).
Hence, as hypothesized, the results indicate that parents report lower levels of parenting stress and higher levels of parental self-efficacy following their participation in the intervention program compared to their initial (baseline) reports. In addition, as we predicted, parents report that their children's inattention and impulsivity/hyperactivity symptoms have decreased since enrolling in the intervention program.
3.3. Feasibility and Acceptability Outcomes
As previously stated, 14 mothers and 1 father with a mean age of 40.1 years old participated in the psychoeducational parent training. Over the course of the 8-week training, the participants attended on average 6.2 sessions (std = .77). In the final session, participants filled out the EPT-F questionnaire to rate their experience with the parent training. The descriptive statistics of the three scores that are extracted from the EPT-F (Total Satisfaction Score, Program’s Usefulness Score, and Trainers’ Assessment Score) and their correlation coefficients can be found in
Table 3. It is worth mentioning that there were not any indications of correlations between the three scores of the evaluation of the program (EPT-F) and the dependent variables of our study.
In general, participants were largely satisfied with the program (mean = 40.60, std = 2.35). More specifically, the majority of parents (80% agree, 13.3% strongly agree) reported that the skills they had learned in the program helped to mitigate their children’s behavior problems and difficulties, while only one parent was not sure (neither agree nor disagree). Furthermore, they all stated (60% agree, 40% strongly agree) that they feel more confident in their ability to use the newly acquired skills to manage any future behavioral problem at home. When asked if their relationship with their child had improved since they began the program, the majority (73.3% agree, 6.7% strongly agree) agreed, while three of them neither agreed nor disagreed. Additionally, they all consented that the program met their expectations (13.3% agree, 86.7% strongly agree) and they would recommend it to parents with related issues (6.7% agree, 93.3% strongly agree).
As far as the usefulness and feasibility of the program’s structure are concerned, participants found the program useful (mean = 63.80, std = 4.24). More precisely, they all stated that the information offered in the program was understandable (26.7% agree, 73.3% strongly agree) and beneficial (20% agree, 80% strongly agree). Furthermore, the majority of parents reported that using the newly acquired skills at home was simple (53.3% agree, 13.3% strongly agree, 33.3% neither agree nor disagree) and useful (46.7% agree, 46.7% strongly agree, 6.7% neither agree nor disagree). Even though some participants (2 parents) expressed their desire for more sessions, the majority thought that the number of sessions was appropriate (40% agree, 46,7% strongly agree). Moreover, all parents reported that the content of the handouts given to them was concise and helpful and that they learned skills they intend to apply in the future.
Regarding the evaluation of the facilitators (trainers) of the program, the participants highly rated their instructors’ competence, preparation, and effectiveness during the sessions.
Taking into consideration the correlations between the three scores extracted from the EPT-F questionnaire, a high positive correlation between the parents’ perceived usefulness of the program and their overall satisfaction (r = .735, p < .01) is revealed. Also, there is a positive correlation between the evaluation of the trainers with both their overall satisfaction from the program (r = .560, p < .05) and their perceived usefulness of the program (r = 646, p < .01).
In addition to the quantitative data, the comments of the participants (qualitative feedback) for the parent training highlighted that they found the program of good quality and providing them with useful information about attention deficits and the co-occurring difficulties observed in children with ADHD. Also, the program contributed to organizing and comprehending their prior knowledge. More importantly, they acknowledged considerable progress on parenting skills and the acquisition of new techniques to interact and support their children effectively. They also pointed out that the psychoeducational training gave them the space to share experiences with other parents, which resulted in empowering each other to endure the adversities. Some of the parents’ insightful comments included the following: “The main benefit I got [from the program] is that I apply better techniques to my child to help him improve a lot.”, “...I gained new knowledge and I shared similar experiences with parents of children facing related adversities.”, “The truth is that I see quite a big change in my child.”, “I am more capable of managing my child.”, “Definitely, I became a more effective parent and was helped to understand my child better”.