Coping with Broken Achievement-Related Expectations in Students from Elementary School: An Experimental Study

Theoretical Background

Expectations are assumptions about what will be or what will happen in the future (Hoorens, 2012). From an early age, people build expectations about their environment and their behavior (e.g., Stahl & Feigenson, 2017). Expectations are useful and important because they can determine future behavior. Wigfield and Eccles (2000) pointed out that expectation of success influences performance, effort, and persistence of behavior. Most studies on children’s and adolescents’ expectations relate to achievement, which is important in the context of school (Pinquart & Ebeling, 2020). The present study addresses expectations of elementary school children about future performance in an attention task. The study assesses how children react to feedback that is discrepant to their expectation.

In order to create expectations about future performance, it is necessary to evaluate one’s own abilities and the challenges of the tasks. This requires cognitive skills that develop during childhood and adolescence. Many preschoolers and children of early elementary school age trend to have overly positive expectations about their future performance, even after receiving negative feedback about their present performance, which might be based, in part, on their limited cognitive abilities to differentiate between hopes or aspirations and their chance to succeed (Muenks, Wigfield, & Eccles, 2018). However, we do not yet know much about how these overly positive expectations change across childhood. For example, Parsons and Ruble (1977) found that three- to five-year-olds have more positive expectations after receiving negative performance-related feedback than 6- to 8-year-olds and 9- to 11-year-olds, while expectations of latter two groups did not differ. Stipek and Hoffman (1980a) observed that 5- to 6-year-old girls had higher expectations after the receipt of negative feedback than 7-8-year-olds but no such difference emerged in boys. In another study from these authors, low-achieving 3rd-graders adjusted their expectations for future achievement even less than low-achieving 1st-graders did, when compared to their high-achieving peers (Stipek & Hoffman, 1980b). The sources of these inconsistent results have not yet been well understood.

As many younger children tend to maintain their positive achievement expectations despite discrepant feedback, knowledge is needed on how they process expectation-discrepant information and protect their expectation against contradictory evidence. A theoretical model by Brandtstädter and colleagues on coping with information that is discrepant to the present self-concept (e.g., Brandtstädter 2007; Brandstädter & Greve, 1994) has been expanded to coping with expectation violation (the VIOLEX model; Rief, Glombiewski, Gollwitzer, Schubö, Schwarting, & Thorwart, 2015). The model distinguishes three ways of dealing with discrepant information: First, people could react with active efforts to avoid future discrepant information (assimilative coping). For example, students may work harder for the next class test. Since not all is/ought discrepancies can be solved in this way, there have to be more options than to assimilate. Accommodation is the second coping strategy for adapting to discrepant feedback. For example, after an unexpected bad grade in mathematics, students may decrease their expectation about their future achievement in mathematics. The third way of reacting to an expectation violation is a defensive reaction (immunization) against the unexpected outcome. The so-called data-oriented immunization refers to ignoring or downplaying the discrepant information, such as judging the information as atypical or its source as dubious (Andrew & Hauser, 2011). For example, if the unexpected bad grade at school is ignored or perceived as an exception, there would be no need for assimilation or accommodation. There is another type of immunization—concept-oriented immunization which refers to the reframing of the meaning of a broken expectation so that the former discrepant information is no longer relevant (Greve & Wentura, 2010; Rief et al., 2015). For example, students may now expect to learn new things in mathematics—an expectation that could not be disconfirmed by a bad grade. As we focuses on quantitative rather than qualitative change in expectations, we do not address concept-oriented immunization. While the process of assimilation can be characterized as conscious process (deciding to become active), Brandtstädter characterized accommodation and immunization as rather nonintentional process which is hardly accessible to direct reflection. Nonetheless, the results of these processes (such as non-accuracy of remembered discrepant feedback) are certainly consciously accessible (e.g., Brandtstädter, 2007).

Scientific knowledge on how children deal with expectation-discrepant information is limited, although some studies on children’s and adolescents’ accommodative coping with discrepant information have been recently published (Lessing, Thomsen, Mähler, & Greve, 2017; Thomsen, 2016). Persistence of overly optimistic expectations in younger children could be based on data-oriented immunization, such as denial and distraction, as immunization seems to be a simple cognitive strategy that does not require complex reframing of a situation (Thomsen, 2016). There is some evidence that the use of data-oriented immunization (Piekney et al., 2017; Thomsen, 2016) in general, and of avoidance in particular (Zimmer-Gembeck & Skinner, 2011), decline from middle childhood to adolescence. However, the results are, to some degree, inconsistent across studies. In principle, children could also use assimilative coping for dealing with an unexpected failure, such as trying harder to succeed. A review of older studies by Compas, Banez, Malcarne, and Worsham (1991) indicated that assimilative coping (“primary control coping”) emerges around the age four or five and increases until late childhood. However, some later studies did not find significant age differences in assimilative coping from late childhood to late adolescence (Fear et al., 2009; Pfeiffer & Pinquart, 2011). Thus, it is not clear whether an age-associated increase in assimilative coping from middle to late childhood would contribute to the tendency of achievement-related expectations becoming more realistic during this age interval (Muenks et al., 2018; Parsons & Ruble, 1977; Stipek & Hofman, 1980a). Finally, the increase in realism of achievement-related expectations from middle to late childhood might also be explained by a higher use of accommodation. In fact, Heckhausen and Schulz (1995) found an increase of flexible goal adjustment during childhood and adolescence, which might be generalized to adjustment of expectations. However, a review by Zimmer-Gembeck and Skinner (2011) on coping strategies associated with flexible adjustment of goals or expectations did not detect a general developmental trend, and the results of the individual studies were highly inconsistent. No age differences in accommodative coping have been found in the preschool years (Lessing et al., 2017) and adolescent years (Thomsen & Greve, 2013, Pfeiffer & Pinquart, 2011). Thus, it is unclear whether an increase in accommodative coping could explain the shift in achievement-related expectations during the elementary school years. In addition, none of the available studies have empirically related age differences in assimilation, accommodation, and immunization to age differences in expectation change after the receipt of discrepant information.

Several researchers have suggested that age differences and interindividual differences in the persistence of achievement-related expectations and in the use of accommodative cooping may be based on related differences in cognitive abilities and executive function (e.g., Muenks et al., 2018; Parsons & Ruble, 1977; Thomsen & Greve, 2013).

Accommodation and, in part, immunization involve shifting perceptions or evaluations of situations or events from a negative to a positive light (or vice versa). Thus, cognitive abilities – such as attention shifting, divergent thinking, inhibition of conflicting responses, and perspective taking – may be relevant for the development of accommodative strategies (Greve & Thompsen, 2014; Thomsen & Greve, 2013). These cognitive preconditions develop across childhood and even in adolescence (Davidson, Amso, Anderson, & Diamond, 2006; Vuontela et al., 2013). In fact, Greve and Thompsen (2013) found a significant, albeit small, positive correlation between accommodative coping and cognitive flexibility (reframing ability) in early adolescence.

Research Questions

We chose a combination of an explorative and confirmatory approach because theory-based, directed assumptions could only be stated for some of our research questions. The first aim of the present study is to investigate the relationship between age and persistence/change of expectations after the receipt of discrepant feedback during the elementary school years. As Parsons and Ruble (1977) and Stipek and Hoffman (1980a) found that younger children were less likely to downgrade their expectations after getting negative achievement-related feedback, our first research question asked whether this result can be replicated in the present study. Based on the VIOLEX model (Rief et al., 2015), the second research question asked whether immunization against expectation-discrepant information, as well as assimilation, would be associated with less of a decrease in expectations after poor achievement, whereas reported accommodation would be related to a greater decrease in expectation.

As age differences in the change of expectations could be based on related differences in the use of assimilative, accommodative, and immunizing coping with discrepant information, the third research question asked whether the levels of immunization, assimilation, and accommodation vary by age, and whether age differences in these strategies explain age differences in expectation change. Finally, the fourth research question asked whether higher cognitive flexibility is associated with higher self-reported use of accommodative coping.

Method

Results

While the participants achieved, on average, 47.90 hits in the first round, the achievement dropped to 32.51 in the second round (a 32.13% decrease), thus indicating that the manipulation worked (Table 1). Children expected a higher achievement in the second round (M = 50.02) compared to their past achievement in the first round (t(114) = 3.87, p < 0.001). After getting the feedback on their low achievement in round two, their expectation for the achievement in the next trial dropped down to 35.75 and 35.80 hits in the two assessments (t(114) = 18.80, p < 0.001 and t(114) = 18.75, p < 0.001). Nonetheless, their expectation for achievement in the third round was higher than their achievement in the second round (t(114) = 4.50, p < 0.001 and t(114) = 4.47, p < 0.001, respectively). During the time where participants were asked to express their thoughts, about half of the children gave comments that could be coded according to the VIOLEX model, with immunization being reported most often (M = 0.48). The other children only verbalized negative feelings about their broken expectations.

Older children had higher achievements in the attention game (r = 0.24, p < 0.01), higher initial expectations about their achievement (r = 0.21, p < 0.05), and higher scores on the cognitive flexibility (reframing; r = 0.19, p < 0.05) (Table 2). Achievement scores showed strong correlational stability (r = 0.62, p < 0.001); expectation scores were also strongly correlated across time (r = 0.52 – 0.88, p < 0.001). We found no significant correlations of the three coping variables (immunization, assimilation, accommodation) with the other study variables. Finally, a better achievement in the attention task was associated with a higher score in the test for cognitive flexibility (r = 0.24, p < 0.01).

In order to test whether change in expectations differed by age, we computed an analysis of variance with repeated measurement with age as the between-subject factor and time of measurement as the within-subject factor. We found a significant main effect of time of measurement (F(2,111) = 305.80, p < 0.001), indicating a decline in expectations. While the main effect of age was not statistically significant (F(3,111) = 1.42, p < 0.25), there was a significant quadratic interaction effect of age and time of measurement (F(3,111) = 4.53, p < 0.005). Figure 1 indicates that, between the first and second assessment, expectations of younger children showed a stronger decline than expectations of older children. While younger children slightly increased their expectations after thinking aloud about their past performance, older children showed a small further decline in their expectations between the second and third assessment. Separate analyses of change between the first two times of measurement and between the second and third time of measurement showed that the interaction effects of age group and time were significant at both time intervals (F(1,111) = 2.95, p < 0.04 and F(1,111) = 3.02, p < 0.035). The main effect of time of measurement was restricted to the first time interval F(1,111) = 360.55, p < 0.001).

OPEN IN VIEWER

Figure 1

Age differences in expectation change.

For answering the second research question, we computed two regression analyses with expectations at the second and third assessment, as dependent variables and expectations at t – 1, as well as reported immunization, assimilation, and accommodation as independent variables. As shown in Table 3, the three coping strategies did not predict change in expectations between the first and second assessment. However, immunization predicted expectation change between the second and third assessment (β= 0.10, t = 2.22, p < 0.05). Children whose verbal reports indicated immunization against the negative achievement-related feedback tended to increase their expectations between the second and third assessment (M = 1.07, SD = 5.21, k = 43), while children who did not report immunization showed a small, further reduction of their expectations (M = –0.56, SD = 3.93, k = 72).

The third research question asked whether the reported levels of immunization, assimilation, and accommodation vary by age. We did not find significant age differences in these three forms of coping with unmet expectations (r = 0.01 to 0.12, p > 0.05; Table 2). Given this finding, there was no need to test whether age differences in expectation change may be based on age differences in immunization, assimilation, and accommodation.

With regard to our last research question, we did not find that cognitive flexibility was a significant predictor of self-reported accommodative and assimilative coping (r = –0.07 and 0.07, p > 0.05; Table 2).

Discussion

The present study assessed changes in expectations of children from elementary school after being confronted with discrepant achievement-related feedback. Younger children showed stronger immediate declines of their expectations followed by stronger increases after having some time to activate coping strategies. Immunization against the expectation-discrepant feedback was followed by a small increase in the expectations while the non-users reported a further decline in expectations. Immunization, assimilation, and accommodation did not vary by participant age or cognitive flexibility. Children tended to have higher expectations about their future achievement as compared to their level of past achievement, indicating some future optimism or the experience that practicing one’s skills (in previous trials) promotes future achievement (Brinkmann, 2012).

Results of previous studies on age differences in children’s expectation change have been inconsistent. The present study indicates that age differences depend, at least in part, on the timing of the assessment of expectations. We observed that younger children downgraded their expectations more immediately after the receipt of negative feedback about their performance, but were more likely to increase their expectations after giving them some time to think about reasons and excuses. Although the immediate accommodative regulation is functional for protecting the self-esteem (by avoiding a following disappointment), lowering expectations also has long-term costs due to anticipating a less positive future, lower striving for success, and possibly lower future success (Putwain, Nicholson, Pekrun, Becker, & Symes, 2019). This dynamic of short-term and long-term consequences of accommodation can explain the nonlinear expectation change in the younger students.

We found age differences only when analyzing expectation change, rather than the absolute levels of expectations at an individual time of measurement. Our results indicate that younger children from elementary school may need more time than older children for activating processes that protect their expectations (immunization or assimilation), which leads to stronger immediate adjustments of expectations after getting discrepant feedback when protective processes do not yet work. In fact, research on emotion regulation indicates that younger children tend to lack the self-control that is required to downregulate and hold back an intense, immediate response to frustration (Jeronimus & Laceulle, 2017). After having some time to think about the feedback, protective processes also come into play with the younger children, thus leading to an attenuation of the previous expectation change and a decline in age differences in expectations. Thus, age differences in children’s facets of the VIOLEX model may mainly be found immediately after the receipt of discrepant feedback. As we assessed the three coping strategies only once, further studies are needed for directly testing the assumption about a delayed activation of coping strategies in younger children.

A larger number of children did not report any strategy of the VIOLEX model but only signs of psychological distress, which indicates that they had not (yet) found a way of coping with their broken expectation. The verbal coping responses of the children mainly indicated an immunization against the discrepant feedback, and only immunization predicted the direction of following expectation change. Assimilative or accommodative strategies may be more difficult to use than data-oriented immunization, such as doubting or ignoring the discrepant feedback. For example, in the case of assimilation, children would have to try harder in the next round, which might not be easy if they already did their best in the previous round. Accommodation requires cognitive flexibility, such as reframing and perspective taking, which is not yet completely developed in the elementary school years (Davidson et al., 2006; Vuontela et al., 2013). However, we cannot rule out an alternative explanation: As immunization referred to doubts about the validity of the discrepant feedback, the prevalence and statistical effect of immunization could also be based on the perception of hints that the feedback was, in fact, manipulated (e.g., the impression that there were too few target animals).

The observed effect of immunization on change in expectations between the second and third assessment was in line with the VIOLEX model, although the effect size was statistically small (Brandtstädter, 2007; Rief et al., 2015). Children do not need to downgrade their expectations after negative feedback if they do not believe in the validity of the feedback or ignore it altogether. The lack of immunization effect on change in expectations between the first and second assessment indicates that immunization did not work as a stable disposition across different time intervals, but had only a situational effect after being activated. Not finding statistical effects of accommodation and assimilation on expectation change should be interpreted with caution as these strategies were rarely mentioned, thus restricting the variance and the chance of finding statistically significant effects.

While our measure of cognitive flexibility correlated with achievement scores in the attention game, it did not correlate with self-reported accommodation and with children’s expectations after the receipt of negative feedback. The contents of the test of cognitive flexibility (Greve & Thomsen, 2013) might have been too unspecific for predicting accommodative tendencies after the receipt of expectation-discrepant feedback. The lack of significant associations may have also been based on variance restriction of the accommodation measure, as this behavior was rarely mentioned.

Limitations and Conclusions

A few limitations of the present study have to be mentioned: First, expectation violation was operationalized via lower than expected achievement. Although this kind of discrepancy between expectation and achievement is most often found in the academic context (Pinquart & Ebeling, 2020), it would also be relevant to test how children respond to a higher-than-expected achievement. Second, as the first assessed expectation was only based on the achievement in one previous trial, it was probably easier to change the expectation in that instance than in other cases where they are based on multiple previous experiences. Third, because the assessment of expectations required basic numeric skills, we did not assess preschoolers. More and/or stronger age differences may appear when expanding the age range of the participants. Fourth, the three strategies of coping with expectation-discrepant information were only assessed via open questions. There is still no valid and reliable multi-item self-report measure that assesses the three coping strategies of the VIOLEX model in elementary school children. Although parental reports on their child’s behavior may be used, they would only assess dispositional tendencies rather than the child’s actual use of strategies when being confronted with discrepant information. A study by Lessing, Mähler, and Greve (2015) found only weak correlations of parental reports on their child’s dispositional flexibility of goal adjustment with a behavioral measure of children’s actual flexible goal adjustment. Fifth, it was difficult to find relevant measures of cognitive flexibility for the assessed age range. Sixth, results may differ if the experimental situation provides different opportunities for the use of a particular strategy, such as opportunities to influence future achievement feedback. Finally, in order to keep the time for the assessment limited, we did not assess a number of possible correlates of expectations and achievement, such as self-efficacy beliefs.

Despite these limitations, the following conclusions can be drawn from the present study: First, researchers who are interested in age differences in expectation change after discrepant feedback across childhood should assess immediate as well as delayed responses because the size and direction of change may vary between these time intervals. Second, as we and previous studies (Lessing et al., 2017; Pfeiffer & Pinquart, 2011; Thomsen & Greve, 2013) found no age differences in the mean levels of immunization, assimilation, or accommodation across childhood and adolescence, future research should shift the focus to other aspects of these strategies, such as how fast these strategies are activated or whether age groups differ in the quality of the individual strategy (e.g., simply ignoring versus reinterpreting discrepant information). As the present pilot study assessed the three strategies of the VIOLEX model in a limited age range (7 –11 years), more research is recommended that expands this age range and optimizes the assessment of these strategies, for example by combining data from more than one measure. Finally, experimental studies on coping with broken expectations should be accompanied with non-experimental studies in order to test the generalizability of the experimental findings in daily life.

Declaration of interest

No authors have any financial and personal relationships with other people or organizations that could have inappropriately influenced our work.

Bio Sketches

Martin Pinquart, Ph.D., Professor of Developmental Psychology at Philipps University in Marburg, Germany since 2007. Research on child and adolescent development as well as gerontopsychology. Member of the Principal Investigators of the Research Training Group (RTG) “Expectation Maintenance vs. Change in the Context of Expectation Violations: Connecting Different Approaches” (DFG Ref. no.: GRK 2271).

Helena Block, M.Sc., worked from 2017–2019 as doctoral student in the RTG “Expectation Maintenance vs. Change in the Context of Expectation Violations: Connecting Different Approaches” at the Department of Psychology at Philipps University of Marburg.