Homework was defined by Cooper (1989) some years ago as the tasks assigned by teachers to students to be completed outside the class. Epstein and van Voorhis (2012) identified homework as a natural connector of school and home. In these ways, homework is one of the most common school activities involving teachers, students, and parents (Rosário et al., 2015). Recently, however, there have been serious debates in Spanish schools and in other countries about whether or not teachers should assign homework. The debates involve students’ complaints about the time required to do their homework, parents’ complaints about the quantity of homework assigned and their lack of information on how to guide their child on homework tasks, and, teachers’ complaints about the lack of time to design effective homework assignments and deliver feedback to students, and the lack of parental support for students to do their work (Cooper et al., 2006).
There are several connections of students’ homework, parental involvement, and student achievement that must be understood to address questions about the value of homework and improving the homework process. These relationships have been frequently studied across the decades, with most studies confirming a positive impact of homework on student achievement (Rosário et al., 2009; Bembenutty and White, 2013). However, findings vary depending on the research design (Cooper et al., 2006; Patall et al., 2008), nature of measures (i.e., global vs. specific) (Trautwein et al., 2009), students’ grade level (Núñez et al., 2015), and focus of the analysis (e.g., student variables, instructional process variables, or parental involvement) (Núñez et al., 2014). Other studies explored the influence of parental involvement on students’ homework behaviors and resulting achievement (Cooper et al., 2001, 2006; Patall et al., 2008; van Voorhis, 2011; Bardou et al., 2012; Dumont et al., 2012; Kim and Fong, 2013).
A substantial number of studies analyzed the association of different student homework behaviors with students’ academic achievement (Xu, 2010; Núñez et al., 2013b; Xu et al., 2014). However, few studies have explored whether and how students’ achievement levels affect their homework behaviors. This study aims to increase understanding on how students’ levels of achievement are related to their homework behaviors (i.e., homework time spent, homework time management, and amount of homework completed), and how students with different achievement levels perceive the involvement of their parents in the homework process (i.e., control and support).
Why Are Parents Involved in Their Children’s Homework?
Relationships between parental involvement in homework and academic achievement have been deeply debated and frequently investigated, with inconsistent results (Gonida and Vauras, 2014). Some studies found a positive relationship (Cooper et al., 2001; Pomerantz and Eaton, 2001), others reported a negative relationship between the two variables (Schultz, 1999). Dumont et al. (2012) found both positive and negative relationships, depending on the nature or quality of the involvement. For example, whereas perceived parent–child conflicts about homework were negatively associated with educational outcomes, perceived parental competence and support for students’ self-direction were positively related to achievement. Similar results were obtained by Karbach et al. (2013), who found that academic achievement was significantly and negatively associated with parental control and strict structure (i.e., excessive control and pressure on children to complete assignments, consistent guidelines and rules about homework and school work).
In a recent study, Núñez et al. (2015) found that students’ perceptions of strong control by parents in the homework process was directly and negatively related to academic achievement. The higher the perceived parental homework control, the lower the students’ academic achievement. In the same study, perceived parental homework support was positively related to the achievement of junior high and high school students, but not to that of elementary school students.
Why do parents become involved in children’s homework? The literature suggests several reasons for parents’ involvement: their own motivation (Hoover-Dempsey et al., 1995; Katz et al., 2011); their socioeconomic status (Davis-Kean, 2005); teacher outreach and homework design that encourages engagement (Hoover-Dempsey and Sandler, 1997; Epstein and Van Voorhis, 2001); and their children’s academic functioning (Pomerantz and Eaton, 2001; Grolnick et al., 2002; Cunha et al., 2015), with academic functioning one of the strongest instigators of parents’ attention to homework.
That is, parents are more likely to be involved when children are not doing well in school (Levin et al., 1997; Pomerantz and Eaton, 2001; Ng et al., 2004; Silinskas et al., 2010). In that situation, parents are more prone to display controlling forms of involvement (Pomerantz and Eaton, 2001; Grolnick et al., 2002; Ng et al., 2004; Niggli et al., 2007). Thus, although a major assumption in previous studies has been that different types of parental involvement in homework are related to different levels of school achievement, it is also likely that children’s academic achievement predicts or motivates parents to become involved in homework in particular ways.
Purpose of This Study
Some studies found that parents’ participation in their children’s academic life (e.g., monitoring progress through conversations with teachers, attending to subjects their children are struggling with) is related to students’ homework completion (Pomerantz et al., 2007) and academic achievement (Wilder, 2013). However, investigations of parental involvement in homework is inconclusive (Patall et al., 2008; Wilder, 2013). Although some authors defend parents’ involvement as a positive practice that can enhance children’s academic success, others describe this support as a time-consuming exercise that frequently generates discomfort, anxiety and conflict in the family (Cooper et al., 2001; Pomerantz et al., 2005a; Patall et al., 2008). However, the majority of findings confirm a positive association between children’s academic functioning (i.e., student achievement and productive homework variables) and parents’ involvement in homework.
Most research has focused on how the context (e.g., family or school) or homework variables (e.g., quantity and quality of homework assignments, parental involvement, students’ homework behaviors) influences student achievement (Trautwein et al., 2002; Hill et al., 2004; Cooper et al., 2006; Pomerantz et al., 2007; Trautwein, 2007; Trautwein et al., 2009; Zhu and Leung, 2012; Karbach et al., 2013; Núñez et al., 2013b, 2015). Few studies, however, flipped the coin to examine the inverse relationship. As Nurmi and Silinskas (2014, p. 455) point out, there is a need to analyze findings from a ‘child-directed development’ perspective, in their own words, “to see that children are not only the passive targets of their parents’ behavior, guidance, and parenting practices but they also influence their parents in many ways.” For example, Chen and Stevenson (1989) and Levin et al. (1997) concluded that when children showed low academic skills, their parents were more likely to monitor the amount and quality of their homework. More recently, Silinskas et al. (2010, 2013) analyzed the behavior of first and second grade students. They reported that the lower the children’s literacy and numeracy, the higher the levels of homework help and monitoring displayed by their parents. Silinskas et al. (2013) reinforced these findings, reporting that children’s achievement had an “evocative impact” on their parents’ behavior. Dumont et al. (2013) analyzed the relationship between fifth and seventh graders’ functioning on homework and the quality of their parents’ homework involvement (conceptualized as a multidimensional construct). They concluded that students’ skills (e.g., levels of reading achievement, reading effort, and homework procrastination) predicted the quality of parental involvement in homework (parental control, parental responsiveness, and parental structure).
This study addresses how children’s levels of prior academic achievement affect their perceptions of whether and how their parents are involved in homework. As in previous studies (Núñez et al., 2015), the dimensions of parental involvement in homework are control (i.e., parents’ pressure on children to complete assignments) and support (i.e., the value students’ place on parents’ assistance and the spirit of parents’ help to support students’ self-direction or autonomy on homework). We explore whether and how student achievement and homework behaviors (i.e., time spent on homework, quality of homework management, and quantity of homework completed) promote specific kinds of parental involvement in homework.
Recent studies (Dumont et al., 2013; Silinskas et al., 2013) using longitudinal designs analyzed the effects of children’s achievement on subsequent parental involvement. In both studies, the direct relationship between the two constructs was estimated with similar results. The associations were negative, indicating that the lower the children’s achievement, the greater the involvement of their parents. However, Dumont et al. (2013) found the significant negative connection only for low achievement on greater parental control, but no significant connection with parental support. By contrast, Silinskas et al. (2013) reported a significant negative effect of children’s reading achievement on both parental monitoring (similar to the Dumont et al., 2013 study) and an even greater or stronger negative effect of achievement on parental support (measured as “parental help”) which (Dumont et al., 2013) did not find in their study. The different findings by Dumont et al. (2013) and Silinskas et al. (2013) may be due to the different ages of the participating students (grades 1 and 2 vs. grades 5 and 7, respectively).
Taken together, the data from these studies indicate that in the early elementary grades, students with low achievement prompted parents’ control and support behaviors, whereas at the junior high school level, students’ low achievement prompted significantly greater control by the parents who were involved. In order to extend analyses on how the level of students’ prior achievement affects their parents’ involvement, this study included measures of the students’ homework behaviors as potential mediating variables as described by Dumont et al. (2013) and Silinskas et al. (2013). Prior studies were not conducted with students or parents at the high school level.
For this study of middle and high school students, a structural equation model (SEM) for homework was elaborated and fitted with the following hypotheses (see Figure 1):
FIGURE 1. Children’s academic functioning and parental involvement relationship.
(1) Prior academic achievement is positively and significantly associated with children’s homework behaviors (i.e., time spent on homework, homework time management, and amount of homework completed);
(2) Children’s homework behaviors are associated with their perceptions of parental involvement in homework;
(3) Prior academic achievement is associated directly and negatively with students’ reports of parental involvement in homework (control and support); and
(4) Perceived parental control and perceived parental support are significantly related.
Previous studies identified grade level as a relevant variable when analyzing the relationships between of academic achievement, students’ homework behaviors and parental homework involvement (e.g., Patall et al., 2008; Skaliotis, 2009; Gonida and Cortina, 2014; Núñez et al., 2015). Thus, in this study the sample was divided into two subgroups (7th and 8th = grades—middle school and 9th and 10th grades—high school) to test the model invariance.
Materials and Methods
A total of 1250 Spanish students from 7th to 10th grade with ages ranging from 12 to 16 years participated in this study. These students attended 68 classes in four urban public schools selected at random from all public schools in Asturias. There were 370 students in grade 7, 346 in grade 8, 257 in grade 9, and 277 in grade 10. Fifty one percent of the participants were male. In the Spanish educational system, compulsory secondary education extends through 9th grade. On average, the families of these students were in the middle class, evidenced by the low percentage of students receiving free or reduced-price lunch (18.7%) as reported in schools’ office data.
Variables and Measures
Students’ perceptions of parental involvement in homework and students’ reports of their own homework behaviors were gathered in questionnaires administered during one regular class period for about 25 min. Students’ prior academic achievement data (report card grades) was provided by the secretary of each school.
Secondary students in middle and high schools are the main actors in their own education, thus students’ reports about their homework behavior and their perceptions of parental involvement provide important views of the homework process. Teachers’ and parents’ actions and messages must be accepted, understood, and processed by the students, themselves, to motivate learning and promote achievement in school (Bempechat, 2004; Epstein, 2011). It is likely, as this study hypothesizes, that the characteristics of students affect how their parents react to them. The data from students provide a good starting place for understanding the research questions in this study.
Parental Involvement in Homework
Two dimensions of parental involvement in homework were assessed: students’ perceptions of control exercised by parents and students’ perceptions of support provided by their parents. The items were adapted from prior studies (e.g., Carter and Wojtkiewicz, 2000; Trautwein and Lüdtke, 2009; Dumont et al., 2012).
Students’ perceptions of parental control were assessed with five items (α = 0.82) (e.g., “My parents are fully aware of me completing all my tasks.”) on a Likert scale with five responses ranging from 1 (completely false) to 5 (completely true). The five items were used to create a latent variable (Parental Control) for the SEM analysis.
Students’ perception of parental support was computed from student responses to three items (α = 0.80) (e.g., “When I have to do homework, explanations by my parents are very useful.”) using the same scoring system as for parental control. A latent variable (Parental Support) was built from the three items for the SEM.
Student Homework Behaviors
Variables of homework behaviors were selected from a pool of items used in other studies (e.g., Núñez et al., 2013b, 2015) to create the latent variables.
Time spent on homework was calculated from student responses to two items (α = 0.70): “How much time do you usually spend on homework each day, Monday through Friday?” and “How much time do you usually spend doing homework during the weekend?” The items were scored on a five point Likert scale, ranging from 1 (less than 30 min), 2 (30 min to 1 h), 3 (1 h to hour and a half), 4 (1 h and a half to 2 h), to 5 (more than 2 h.
Homework time management was calculated from student responses to two items (α = 0.72): “When I’m doing my homework, I get distracted by anything that is around me,” and “When I start homework, I concentrate and do not think about anything else until I finish (reverse coded).” These items were rated on a five-point Likert scale, ranging from 1 (always) to 5 (never).
Amount of homework completed was assessed from student responses to the following question: “Usually, how many tasks do you complete from the assigned homework?” This item was rated on a five-point Likert scale, ranging from 1 (none) to 5 (all).
Prior Academic Achievement
Prior academic achievement was obtained from students’ report card grades in mathematics, Spanish language, English language, and social sciences at the end of the academic year (June) (see Núñez et al., 2015). The grades for the four subjects were used to build a latent variable (Prior Academic Achievement). The measurement scale of this variable ranged from 0 to 10 with 5 as a passing grade.
Participating students were volunteers with approval from their parents. Researchers signed agreements with the collaborating school boards to conduct workshops for participating teachers and for parents on the results and educational implications of the research. All measures except prior academic achievement were collected in October at the beginning of the school year. In the current study the measure for prior academic achievement refers to students’ achievement at the end of the previous school year and is used as an explanatory variable.
Data Analysis Strategy
To address the research questions of this study, data were analyzed in several stages. First, we calculated and analyzed descriptive statistics of the variables in the homework model. Second, following Núñez et al. (2015), three models were compared to examine to what extent the students’ homework behaviors mediated the association between students’ prior academic achievement and perceived parental involvement in homework: a full mediation model (M1), a partial mediation model (M2) [M1 plus a direct path from prior academic achievement to perceived parental involvement in homework (control and support)], and a non-mediation model (M3) [only the direct path from prior academic achievement to homework parental involvement] (see Figure 2). Information criteria-based model selection tools were used to compare the fit to the data of the three candidate models, and select the best (see Vallejo et al., 2014).
FIGURE 2. Structural equation model (SEM) of children’s academic functioning and homework parental involvement (full, partial, and non-mediation models). PC1,..., PC5 (measures of perceived Parental Control), PS1,..., PS3 (measures of perceived Parental Support), TS1 and TS2 (measures of Time Spent on HW Completion), TM1 and TM2 (measures of HW Time Management), HWC (measure of Amount of HW Completed), SL (measure of Spanish Language Achievement), Mt (measure of Mathematics Achievement), EL (measure of English Language Achievement), SS (measure of Social Sciences Achievement). V1 to V5 represent the variance explained. X1 to X5 and Y1 to Y12 are measurement errors.
Third, multi-group analyses were conducted to check the invariance of the homework model chosen for the two subgroups of students at the middle and high school levels. Finally, the best-fit model was used to examine the three hypotheses of the study.
To account for the hierarchical structure of the data (i.e., students in classes), the homework model was fitted with Mplus 5.1 (Muthén et al., 1998–2007) using “type = complex” in the analysis command and “cluster = class” in the variable command. This procedure allowed computation of the standard errors and chi-square tests of model fit, taking into account clustering information and/or non-independence of observations, such as adjusting the standard errors of the regression coefficients. The MLR estimator in Mplus 5.1 (maximum likelihood robust) was selected, which is sensitive to non-normality and non-independence of observations.
A series of statistics and indices were used at different stages of data analysis. Akaike’s (1974) Akaike’s information criterion (AIC), Raftery’s (1993), Bayesian information criterion (BIC), and Browne and Cudeck’s (1993), Browne- and Cudeck’s criterion (BCC) were used to select the proper mediation model. Then, to assess the fit of the model chosen, in addition to chi-square (χ2) statistics and their associated probability (p) values, we used two absolute indices, the goodness-of-fit-index (GFI) and the adjusted goodness-of-fit-index (AGFI); a relative index, the Tucker Lewis Index (TLI) and the comparative fit index (CFI) (Bentler, 1990); and a close-fit parsimony-based index, the root mean square error of approximation (RMSEA), and their 90% confidence intervals (Hu and Bentler, 1999). According to these authors, a model fits well when: GFI, AGFI, and TLI > 0.90, CFI > 0.95, and RMSEA ≤ 0.05.
Table 1 shows descriptive statistics and the correlation matrix for the observed variables in the model. The variables are significantly inter-correlated. Because maximum likelihood (ML) can produce biases when variables fail to follow a normal distribution, we examined the distributions of all the variables (i.e., kurtosis and skewness). Taking the criterion of Finney and DiStefano (2006), for whom 2 and 7 are the maximum allowable values for skewness and kurtosis, respectively, all of the variables respected those criteria (see Table 1). Therefore, with normality conditions met, we fitted the model using MLR.
TABLE 1. Descriptive statistics of the variables in the structural equation homework model (n = 1250 middle and high school students).
Selecting the Best Model
The analyses of the comparison models showed that the fit of the non-mediation model was the worst of the three models (see Table 2). By comparison, the partial mediation model and the full mediation model showed a satisfactory fit, with the best fit of all provided by the partial mediation model [Δχ2(2) = 68.23, p < 0.001]. The likelihood ratio test procedure was favorable to the partial mediation model (M2). Also, to select the best fit model, the statistics provided by AIC, BIC, and BCC were used to determine which of the two models (full or partial mediation model) was more likely to accurately describe the relationships in the matrix data.
TABLE 2. Results of homework model comparison strategy.
Table 2 shows that the partial mediation model has a more valid BIC value than does the full mediation model. Similarly, efficient criteria (i.e., AIC) which tends to choose more complex models (Vallejo et al., 2014), as well as consistent criteria (i.e., BIC), which tends to choose simpler models, favored the selection of the Partial Mediation Model (M2). Based on the suggestions by Burnham and Anderson (2002), we selected M2 as the actual Kullback–Leibler best model for the population of possible samples.
Grade Level Invariance Analysis
The hypothesis of the invariance of the Homework Partial Mediation Model in the two samples of students (7th -8th grade vs. 9th -10th grade) was analyzed with multi-group analyses. Specifically, we tested the similarity of the Homework Partial Mediation Model in both samples with regard to its five dimensions: measurement weights, structural weights, structural covariances, structural residuals, and measurement residuals.
Results showed that the hypothesized homework model is similar in both samples on four of the five criteria (see Table 3). Assuming that the unconstrained model is correct [χ2 = 577.614, df = 212, p < 0.001, χ2/df = 2.725, GFI = 0.948, AGFI = 0.924, CFI = 0.957, RMSEA = 0.037, 90% CI (0.034, 0.041), p = 1.000], when testing equality in measurement weights, in structural weights, in structural covariances, and in structural residuals no statistically significant differences were found. Finally, assuming the absence of differences in structural residuals, statistically significant differences were found in measurement residuals.
TABLE 3. Results of grade level invariance analysis.
Therefore, the results show that the Homework Partial Mediation Model is invariant for the two groups of students in the first four dimensions (measurement weights, structural weights, structural covariances, and structural residuals), but not for the last one (measurement residuals). The analysis of structural weights and structural covariances was the main focus of the multi-group analysis, which indicates the appropriateness of using the total sample to analyze the homework model.
Children’s Prior Academic Achievement and Perceived Parental Involvement in Homework
Results for the Homework Partial Mediation Model adjustment are provided in Table 4 and Figure 3. Overall, the analyses confirm the three hypotheses initially established for the study. First, as hypothesized, prior academic achievement was significantly associated with students’ homework behaviors. Statistically significant and positive associations were found between prior academic achievement and the time students spend on homework, the management of this time, and the amount of homework completed.
TABLE 4. Standardized and unstandardized regression weights, standard errors, z-values, and associated p-values for the Homework Partial Mediation Model.
FIGURE 3. Standardized total effects in the Homework Partial Mediation Model (N = 1250). VE (Variance Explained). All coefficients are statistically significant at p < 0.001, except Homework Time Management on Perceived Parental Support (p < 0.01), and Prior Academic Achievement on Perceived Parental Support (p > 0.05, not statiscally significant).
Second, children’s homework variables and perceived parental homework involvement were significantly and positively related: time spent on homework with perceived parental control and with perceived parental support; time homework management with perceived homework parental control and with perceived homework parental support; and, finally, amount of homework completed with perceived homework parental control and with perceived homework parental support. Third, the direct association between prior academic achievement and perceived parental involvement in homework was significant and negative for perceived parental control, but contrary to our hypothesis not statistically significant for perceived parental support. It is interesting to note, however, that the indirect association between prior performance and perceived homework parental involvement (through time spent on homework, homework time management, and amount of homework completed) was positive and significant for both types of perceived homework parental involvement: support and control. Four, both dimensions of perceived parental involvement in homework were positive and strongly related (r = 0.573, d = 1.40).
Additionally, data indicate that both dimensions of perceived parental homework involvement also were moderately predicted by children’s achievement levels and students’ homework behaviors (see Figure 3): 21.3% (perceived homework parental control) and 14.1% (perceived homework parental support).
Pomerantz et al. (2007, p. 399) suggested that research is needed on how children’s characteristics influence their interactions with and the involvement of parents on school work. In their words: “[the] consideration of the match between children’s characteristics and the manner in which parents become involved is a crucial endeavor.” Their call identified an important research agenda that has not been adequately addressed. Parental involvement does not “produce” student achievement. Rather the parents’ attitudes and actions must flow to and through students, who must interpret and respond to the involvement activities with their own attitudes and actions.
This study responds to that call by focusing on whether student characteristics affect their views of parental involvement. Tests of the data favored SEM analysis of a partial mediation model to explore connections of students’ prior levels of achievement, homework behaviors, and perceived parental involvement in homework. Two major topics emerged that are important to discuss.
Prior Academic Achievement and Students’ Homework Behaviors Predict Perceived Parental Involvement in Homework
This study explored the connections of middle and high school students’ prior levels of achievement and reported homework behaviors with students’ perceptions of the nature of their own parents’ involvement in homework. Findings indicated that level of achievement was related to students’ perceptions of how their parents behaved about homework. Specifically, the data showed that the higher the students’ prior achievement, the more time they spent on their homework, the more homework was completed, and the better their homework time management. Further, the more time spent and the more homework completed, the stronger students’ reports of their parents’ involvement in terms of both control and support of homework.
These findings are aligned with other studies that examined the relationship of these variables in the opposite, more common direction. For example, Núñez et al. (2015), found that the more students’ reported their parents’ involvement, the more time they spent doing homework, the better their time management, and the higher their academic achievement.
By examining different assumptions about the direction of influence of children’s characteristics and parents’ engagement in homework, we can see that, however viewed, students with higher prior achievement tend to spend more time on homework, manage it better, and do more homework. With achievement level taken into account, students who take time to do their homework, perceive and report that their parents continue to offer controlling and supportive messages about the homework process.
This study reinforced prior findings that when secondary school students’ academic performance is poor, they tend to spend less time doing homework, manage their time less effectively, and complete less homework. This study extends prior results by showing that low-achieving students perceived and reported that their parents proffered more controlling messages about homework. Pomerantz et al. (2005b) claimed that children with a history of poor academic performance may be particularly sensitive to the quantity and quality of parental involvement. Low-achieving students, even at the secondary level, may need extra attention from parents to keep them invested in the homework process. If parental involvement is more controlling for these children as suggested in this study and by Núñez et al. (2015), the low-achieving students may progressively disengage from their homework and school tasks. Longitudinal data are needed to examine if parents’ extra pressure helps low achievers improve their achievement scores and stay in school compared to low achievers whose parents ignore the homework process.
The Direct Relationships of Students’ Levels of Prior Academic Performance with Parental Involvement in Homework Differ for Perceived Parental Controlling and Supportive Behaviors
As in prior investigations, this study found a direct negative relationship between children’s academic performance and students’ perceptions of parental involvement in homework, particularly parents’ controlling behaviors. With other variables statistically controlled, students with lower achievement reported that their parents conducted more monitoring and controlling behaviors about homework. There was no significant relationship of student achievement with perceived parental support of homework.
Some researchers explain this pattern of results as reflecting parents’ recognition that low-achieving children need more direction and control than do more successful students, who take more personal responsibility for completing their homework (Pomerantz and Eaton, 2001; Grolnick et al., 2002; Pomerantz et al., 2005a; Epstein and van Voorhis, 2012). Others explain that some parents lack confidence and competence to guide their children in other ways than by controlling (Hoover-Dempsey and Sandler, 1997). Still, others suggest that parents will be more controlling if they and the children have negative attitudes toward homework or behavior problems while doing homework (Fuligni et al., 2002; Pomerantz et al., 2005a), or if parents feel less competent to help children work independently on homework (Pomerantz and Eaton, 2001), or if the child and parents area frustrated by persistent low school achievement (Pomerantz et al., 2005a). As Pomerantz et al. (2007, p. 383) note “when parents’ involvement is controlling, children do not have the experience of solving challenges on their own,” and, “when parents are controlling, they may deprive children of feeling that they are autonomous, effective agents.”
A pattern of over-control by parents may not help students who are struggling to improve their achievement. Several studies reported a connection of high control of homework by parents and children low academic achievement (Cooper et al., 2000; Dumont et al., 2012; Karbach et al., 2013; Núñez et al., 2015). These students may be particularly sensitive to parents’ pressure about homework and may not understand the parents’ intent to motivate them to do their work. The findings from this and other research suggest the existence of a vicious circle in the relationship between children’s prior academic achievement, perceived parental involvement in control of homework, and children’s later achievement. That is, unless parental involvement in homework is carefully balanced with caring control and support messages, low-achieving students may avoid homework and disengage from school, especially in the secondary grades. To break the cycle, this study suggests, an optimal combination of control and support messages is needed to encourage middle and high school students to spend time on, manage, and complete their homework.
A substantial amount of research has analyzed the association of various student homework behaviors (e.g., time spent on homework, time management, amount of homework done, procrastination, emotions, goals and motivations for doing homework, attitudes) with students’ academic achievement. Literature also is replete with studies of how parental involvement in homework affects students’ academic achievement. However, few studies flipped the coin to examine how students’ prior achievement levels affect their homework behaviors and how children’s academic functioning affects parents’ control or support of homework.
This study examined the little known associations of secondary students’ achievement levels with other important elements of the homework process—students’ behaviors and parents’ involvement. The findings indicated that children’s academic functioning was associated with their perceptions of parental involvement in the homework process. The study reveals the recursive nature of these important components of the homework process: children’s achievement level affects perceived parental involvement in homework, and, over time, parental involvement in homework affects students’ later performance.
This study supports and extends the results of past research to support an interactive model of socialization (Collins et al., 2000; Grusec, 2002). The behaviors of parents and children are modeled progressively as their interactions proceed and progress, and as results accumulate to shape the trajectory of student learning.
The presumed reciprocal relationships of parent involvement and student achievement are provocative, but they must be examined in future studies. This study’s data were cross-sectional with one-time measures of the independent and dependent variables. This prohibits claims of causality. Future studies using the required longitudinal data and/or experimental designs that guide specific parental behaviors and messages could test the assumption of recursive relationships of parental involvement and student achievement, which may affect each other, over time.
Another limitation of this study is that all measures were reported by the students. This helped us learn what students with different levels of achievement say about their homework time and products, and how they view the involvement of their parents. Although important, one set of reporters is not sufficient for fully understanding the roles and relationships of students and parents that affect the homework process. Behavior-based measures from students (such as a homework diary) and data from parents of their involvement in homework are needed, along with the children’s views, to study whether multiple reporters explain their behaviors in the same way. Multiple measures from multiple reporters would confirm or challenge the accuracy of reports of students’ homework behaviors of time spent and homework completed, and build a more robust understanding of the complex and continuous influences on student achievement.
Although research on all aspects of the homework process must continue to improve, the results of this and prior studies (e.g., Grolnick and Slowiaczek, 1994; Epstein, 1995; Cooper and Valentine, 2001; Hill and Taylor, 2004; Pomerantz et al., 2005b; Epstein, 2007), have clear and useful educational applications. Numerous studies confirm that, over time, parental involvement with students on homework is associated with higher student achievement. Positive practices of parental involvement may promote students’ cognitive, linguistic, and mathematical skills, metacognitive skills, and strategies for a self-regulated learning, as well as positive attitudes toward school and motivation to learn. However, as Darling and Steinberg (1993) alerted, the effect of these practices is largely determined by the style in which the practices are carried out. And, results also depend on the quality of the design and clarity of parental involvement activities on homework with specific learning goals for students (Epstein and van Voorhis, 2012). The present study and previous research suggest that parental involvement in homework should be more strongly characterized by autonomy support, process focus, positive affect, and positive beliefs in students’ abilities than by too much control, negative affect, and negative beliefs about homework. School administrators, school psychologists, and teachers should offer workshops for secondary school parents on core aspects of their involvement in homework (e.g., how to support students’ independent thinking and completion of assignments; how to prevent and cope with children’s emotional distress about homework; and how to maintain student motivation but reduce undue parental pressure, particularly on students who are struggling in school).
The results of this and other studies suggest that, with teachers’ guidance and materials, more parents could help their students (a) strive to be more independent in their study (Núñez et al., 2013a); (b) understand that their effort (not innate ability) will help them complete their assignments; (c) focus on the positive aspects of school, homework, and learning rather than on negative attitudes (Cunha et al., 2015); and (d) face homework with self-confidence not just to avoid failure but to complete tasks, solve problems, and meet success. The results of this study deepen our understanding about the potential for parents’ positive interactions with their teens on homework at the secondary school level.
This study was carried out in accordance with the recommendations of the ethics committee of the University of Oviedo. All subjects gave written informed consent in accordance with the Declaration of Helsinki.
NS was responsible for the data collection, and for the literature search with JN. GV was responsible for data analysis, and JE for the data interpretation. JN, JE, PR, and AV made important intellectual contribution in research design and manuscript revision. All authors were involved in the writing process of this manuscript.
This work has been funded by the Department of Science and Innovation (Spain) under the National Program for Research, Development and Innovation: project EDU2014-57571-P, and from the European Union, through the European Regional Development Funds and the Principality of Asturias, through its Science, Technology and Innovation Plan (grant GRUPIN14-100 and GRUPIN14-053).
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This manuscript was completed with the help of funding from Ministry of Science and Innovation of Spain (Ref.: EDU2014-57571-P, EDU2013-44062-P, and PSI2011-23395).
Akaike, H. (1974). A New Look at the Statistical Model Identification. IEEE Trans. Automat. Control 19, 716–723. doi: 10.1109/tac.1974.1100705
CrossRef Full Text | Google Scholar
Academic Goals, Student Homework Engagement, and Academic Achievement in Elementary School
Antonio Valle,1,*Bibiana Regueiro,1José C. Núñez,2Susana Rodríguez,1Isabel Piñeiro,1 and Pedro Rosário3
1Department of Developmental and Educational Psychology, University of A Coruña, A Coruña, Spain
2Department of Psychology, University of Oviedo, Oviedo, Spain
3Departmento de Psicologia Aplicada, Universidade do Minho, Braga, Portugal
Edited by: Jesus De La Fuente, University of Almería, Spain
Reviewed by: Melinda J. Mollette, Gwinnett County Public Schools, USA; Javier Fiz Pérez, Università Europea di Roma, Italy
*Correspondence: Antonio Valle se.cdu@rallav
This article was submitted to Educational Psychology, a section of the journal Frontiers in Psychology
Author information ►Article notes ►Copyright and License information ►
Received 2015 Nov 1; Accepted 2016 Mar 15.
Copyright © 2016 Valle, Regueiro, Núñez, Rodríguez, Piñeiro and Rosário.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
This article has been cited by other articles in PMC.
There seems to be a general consensus in the literature that doing homework is beneficial for students. Thus, the current challenge is to examine the process of doing homework to find which variables may help students to complete the homework assigned. To address this goal, a path analysis model was fit. The model hypothesized that the way students engage in homework is explained by the type of academic goals set, and it explains the amount of time spend on homework, the homework time management, and the amount of homework done. Lastly, the amount of homework done is positively related to academic achievement. The model was fit using a sample of 535 Spanish students from the last three courses of elementary school (aged 9 to 13). Findings show that: (a) academic achievement was positively associated with the amount of homework completed, (b) the amount of homework completed was related to the homework time management, (c) homework time management was associated with the approach to homework, (d) and the approach to homework, like the rest of the variables of the model (except for the time spent on homework), was related to the student's academic motivation (i.e., academic goals).
Keywords: homework, academic goals, student homework engagement, approach to homework, academic achievement, elementary school
Literature indicates that doing homework regularly is positively associated with students' academic achievement (Zimmerman and Kitsantas, 2005). Hence, as expected, the amount of homework done is one of the variables that shows a strong and positive relationship with academic achievement (Cooper et al., 2001).
It seems consensual in the literature that doing homework is always beneficial to students, but it is also true that the key for the academic success does not rely on the amount of homework done, but rather on how students engage on homework (Trautwein et al., 2009; Núñez et al., 2015c), and on how homework engagement is related with student motivation (Martin, 2012). There is, therefore, a call to analyze the process of homework rather than just the product; that is, to examine the extent to which the quality of the process of doing homework may be relevant to the final outcome.
Trautwein's model of homework
The model by Trautwein et al. (2006b) is rooted in the motivational theories, namely the theory of the expectancy value (Eccles (Parsons) et al., 1983; Pintrich and De Groot, 1990), and the theory of self-determination (Deci et al., 2002), as well as on theories of learning and instruction (Boekaerts, 1999). Trautwein and colleagues' model analyzes students' related variables in two blocks, as follows: the motivational (aiming at directing and sustaining the behavior) and the cognitive and behavioral implications (cognitions and behaviors related to the moment of doing homework).These two blocks of variables are rooted in the literature. Motivational variables are related with the theory of expectancy-value by Eccles (Parsons) et al. (1983), while the variables addressing students' implication are related with the school engagement framework (e.g., Fredricks et al., 2004). However, as Eccles and Wang (2012) stress, both models are interrelated due to the fact that both variables are closely related and show reciprocal relationships.
Student homework engagement: the interplay between cognitive and behavioral components
Engagement is a relatively new construct with great relevance in the field of psychology and instruction (Fredricks et al., 2004). Generally considered, engagement has been described as the active implication of the person in an activity (Reeve et al., 2004). However, despite the close relation between engagement and motivation, literature clearly differentiates between them (e.g., Martin, 2012), stressing engagement as the behavioral manifestation of motivation (Skinner and Pitzer, 2012), or arguing that motivation is a precursor of engagement rather than part of it. In sum, motivation relates to the “why” whereas the engagement focuses on the “what” of a particular behavior.
Consistent with this perspective, the current research fitted a model with the variable engagement mediating the relationship between motivation and academic achievement (see Eccles and Wang, 2012). Engagement is a complex construct with observational and non-observational aspects (Appleton et al., 2008). Some researchers conceptualize engagement with two dimensions—behavior and emotions (e.g., Marks, 2000)—while others define engagement with four dimensions—academic, behavioral, cognitive, and emotional (e.g., Appleton et al., 2006). In the current study, we followed Fredricks' et al. (2004) conceptualization of engagement as a construct with three dimensions: cognitive (e.g., approaches to learning), behavioral (e.g., student homework behaviors), and emotional (e.g., interest, boredom). For the purpose of the present study, the dimension of emotion was not included in the model (see Figure 1).
General model hypothesized to explain the relationship between academic motivation, student homework engagement, and academic achievement.
Cognitive homework engagement
In the past few decades, a robust body of research has been addressing the relationship between the way students deal with their learning process and academic outcomes (Marton and Säljö, 1976a,b; Struyven et al., 2006; Rosário et al., 2010a, 2013a). Marton and Säljö (1976a,b) examined how students studied an academic text and found two ways of approaching the task: a surface and a deep approach. The surface approach is characterized by learning the contents aiming at achieving goals that are extrinsic to the learning content. In contrast, the deep approach is characterized by an intrinsic interest in the task and students are likely to be focused on understanding the learning content, relating it to prior knowledge and to the surrounding environment (Entwistle, 2009; Rosário et al., 2010b). The metaphor “surface vs. deep” constitutes an easy to perceive conceptual framework, both in the classroom setting and in other educational settings (i.e., doing homework at home), and has been shown to be a powerful tool for parents, teachers, and students when conceptualizing the ways students approach school tasks (Entwistle, 1991; Rosário et al., 2005). The core of the concept of approaches to studying (or to learning) is the metacognitive connection between an intention to approach a task and a strategy to implement it (Rosário et al., 2013b).
The process of doing homework focuses on what students do when completing homework, that is, how they approach their work and how they manage their personal resources and settings while doing homework. It is likely that students' approaches to homework may influence not only the final homework outcome but also the quality of that process. Students who adopt a deep approach are likely to engage their homework with the intention of deepening their understanding of the knowledge learned in class. In this process, students often relate the homework exercises to prior knowledge and monitor their mastery of the content learned. This process involves intrinsic intention to understand the ideas and the use of strategies to build meaning (Cano et al., 2014). In contrast, students who approach homework with a surface approach are likely to do homework with extrinsic motivation (e.g., rewards of their parents, fear of upsetting their teacher). Their goals may target finishing homework as soon as and with the less effort possible to be able to do more interesting activities. Students using this approach are more likely to do homework to fulfill an external obligation (e.g., hand in homework in class and get a grade), than for the benefits for learning.
Behavioral homework engagement
Findings from prior research indicate that the more the implication of students in doing their homework the better the academic achievement (Cooper et al., 2006). Following Trautwein et al. (2006b), our conceptualization of student homework engagement includes behaviors related with the amount of homework done, time spent on homework, and homework time management (e.g., concentration). In the present investigation, these three variables were included in the model (see Figure 1).
Extant findings on the relationship between the amount of homework done and academic achievement are in need of further clarification. Some authors argue for a strong and positive relationship (e.g., Cooper et al., 2006), while others found that this relationship is higher throughout schooling (Cooper et al., 2001; Zimmerman and Kitsantas, 2005). Authors explained this last finding arguing that the load of homework assigned by teachers vary throughout schooling, and also that the cognitive competencies of students are likely to vary with age (Muhlenbruck et al., 2000). More recently, Núñez et al. (2015c) found that the relationship between these two variables varied as a function of the age of the students enrolled. Particularly, this relationship was found to be negative in elementary school, null in junior high school, and positive in high school.
Moreover, the relationship between the amount of homework done and academic achievement relates, among other factors, with the students' age, the quality of the homework assigned, the type of assessment, and the nature of the feedback provided. For example, some students may always complete their homework and get good grades for doing it, which does not mean that these students learn more (Kohn, 2006). In fact, more important than the quantity of the homework done, is the quality of that work (Fernández-Alonso et al., 2014).
Another variable included in the model was the time spent on homework. Findings on the relationship between time spent on homework and academic achievement are mixed. Some studies found a positive relationship (Cooper et al., 2001, 2006) while others found a null or a negative one (Trautwein et al., 2006b, 2009). In 2009, Dettmers, Trautwein and Lüdtke conducted a study with data from the PISA 2003 (Dettmers et al., 2009). Findings on the relationship between the number of hours spent on homework and academic achievement in mathematics show that the students in countries with higher grades spend fewer hours doing homework than students in countries with low academic grades. At the student level, findings showed a negative relationship between time spent on homework and academic achievement in 12 out of 40 countries.
The relationship between the amount of homework done, time dedicated to homework, and academic achievement was hypothesized to be mediated by the homework time management. Xu (2007) was one of the pioneers examining the management of the time spent on homework. Initially, Xu (2007) did not find a relationship between time management and academic achievement (spend more time on homework is not equal to use efficient strategies for time management). Latter, Xu (2010) found a positive relationship between students' grade level, organized environment, and homework time management. More recently, Núñez et al. (2015c) found that effective homework time management affects positively the amount of homework done, and, consequently, academic achievement. This relationship is stronger for elementary students when compared with students in high school.
Academic motivation and student homework engagement relationship
Literature has consistently shown that a deep approach to learning is associated positively with the quality of the learning outcomes (Rosário et al., 2013b; Cano et al., 2014; Vallejo et al., 2014). The adoption of a deep approach to homework depends on many factors, but students self-set goals and their motives for doing homework are among the most critical motivational variables when students decide to engage in homework.
Literature on achievement motivation highlights academic goals as an important line of research (Ng, 2008). In the educational setting, whereas learning goals focus on the comprehension and mastery of the content, performance goals are more focused on achieving a better performance than their colleagues (Pajares et al., 2000; Gaudreau, 2012).
Extant literature reports a positive relationship between adopting learning goals and the use of cognitive and self-regulation strategies (Elliot et al., 1999; Núñez et al., 2013). In fact, students who value learning and show an intention to learn and improve their competences are likely to use deep learning strategies (Suárez et al., 2001; Valle et al., 2003a,b, 2015d), which are aimed at understanding the content in depth. Moreover, these learning-goal oriented students are likely to self-regulate their learning process (Valle et al., 2015a), put on effort to learn, and assume the control of their learning process (Rosário et al., 2016). These students persist much longer when they face difficult and challenging tasks than colleagues pursuing performance goals. The former also use more strategies oriented toward the comprehension of content, are more intrinsically motivated, and feel more enthusiasm about academic work. Some researchers also found positive relationships between learning goals and pro-social behavior (e.g., Inglés et al., 2013).
Reviewing the differentiation between learning goals and performance goals, Elliot and colleagues (Elliot and Church, 1997; Elliot, 1999; Elliot et al., 1999) proposed a three-dimensional framework for academic goals. In addition to learning goals, performance goals were differentiated as follows: (a) performance-approach goals, focused on achieving competence with regard to others; and (b) performance-avoidance goals, aimed at avoiding incompetence with regard to others. Various studies have provided empirical support for this distinction within performance goals (e.g., Wolters et al., 1996; Middleton and Midgley, 1997; Skaalvik, 1997; Rodríguez et al., 2001; Valle et al., 2006). Moreover, some authors proposed a similar differentiation for learning goals (Elliot, 1999). The rationale was as follows: learning goals are characterized by high engagement in academic tasks, so an avoidance tendency in such goals should reflect avoidance of this engagement. Hence, students who pursue a work avoidance goal are likely to avoid challenging tasks and to put on effort to do well, only doing the bare minimum to complete the task. In general, learning goals are associated with a large amount of positive results in diverse motivational, cognitive, and achievement outcomes, whereas performance goals have been linked to less adaptive outcomes, or even to negative outcomes (Valle et al., 2009).
Aims of this study
Several relationships between motivational, cognitive, and behavioral variables involving self-regulated learning in the classroom have recently been studied (Rosário et al., 2013a). However, there is a lack of knowledge of the relationships between these variables throughout the process of doing homework.
The principal purpose of this work (see Figure 1) is to analyze how student homework engagement (cognitive and behavioral) mediates motivation and academic performance. This study aims to provide new information about an issue that is taken for granted, but which, as far as we know, lacks empirical data. The question is: to what extent students acknowledge homework as a good way to acquire competence, improve their skills and performance? Our working hypothesis is that student value homework in this regard. Therefore, we hypothesized that the more students are motivated to learn, the more they will be involved (cognitively and behaviorally) in their homework, and the higher their academic achievement.
To address this goal, we developed a path analysis model (see Figure 1) in which we hypothesized that: (a) the student's motivational level is significantly related to their cognitive homework engagement (i.e., the approach to studying applied to homework), and their behavioral homework engagement (i.e., amount of time spent and homework time management, and amount of homework completed); (b) student's cognitive and behavioral homework engagement are positively associated with academic achievement; and (c) cognitive and behavioral homework engagement are related (the more deep cognitive engagement, the more time spent and time management, and the more amount of homework is done).
The study enrolled 535 students, aged between 9 and 13 (M = 10.32, SD = 0.99), of four public schools, from the last three years of the Spanish Elementary Education (4th, 5th, and 6th grade level), of whom 49.3% were boys. By grade, 40.4% (n = 216) were enrolled in the 4th grade, 35.1% (n = 188) in the 5th grade, and 24.5% (n = 131) in the 6th grade.
The level and type of motivation for academic learning was assessed with the Academic Goals Instrument (Núñez et al., 1997). Although, this instrument allows differentiating a broad range of academic goals, for the purposes of this work, we only used the subscale of learning goals (i.e., competence and control). The instrument is rated on a 5-point Likert-type scale, with responses ranging from one (not at all interested) to five (absolutely interested in learning and acquiring competence and control in the different subjects). An example item is: “I make an effort in my studies because performing the academic tasks allows me to increase my knowledge.” The reliability of the scale is good (α = 0.87).
Approach to homework
To measure the process of approaching homework, we adapted the Students' Approaches to Learning Inventory (Rosário et al., 2010a, 2013a), taking into account both the students' age and the homework contexts. This instrument is based on voluminous literature on approaches to learning (e.g., Biggs et al., 2001; Rosário et al., 2005), and provides information about two ways of approaching homework. For the purpose of this research, we only used the deep approach (e.g., “Before starting homework, I usually decide whether what was taught in class is clear and, if not, I review the lesson before I start”). Students respond to the items on a 5-point Likert-type scale ranging from one (not at all deep approach) to five (completely deep approach). The reliability of the scale is good (α = 0.80).
Time spent on homework, homework time management, and amount of homework completed
To measure these three variables, we used the Homework Survey (e.g., Rosário et al., 2009; Núñez et al., 2015a,b; Valle et al., 2015b,c). To measure the time spent on homework, students responded to three items (in general, in a typical week, on a typical weekend) with the general formulation, “How much time do you usually spend on homework?,” with the response options 1, <30 min; 2, 30 min to 1 h; 3, 1 h to an hour and a half; 4, 1 h and a half to 2 h; 5, more than 2 h. Homework time management was measured through the responses to three items (in general, in a typical week, on a typical weekend) in which they were asked to indicate how they managed the time normally spent doing homework, using the following scale: 1, I waste it completely (I am constantly distracted by anything); 2, I waste it more than I should; 3, regular; 4, I manage it pretty much; 5, I optimize it completely (I concentrate and until I finish, I don't think about anything else). Finally, the amount of homework completed by students (assigned by teachers) was assessed through responses to an item about the amount of homework usually done, using a 5-point Likert-type scale (1, none; 2, some; 3, one half; 4, almost all; 5, all).
Assessment of academic achievement was assessed through students' report card grades in Spanish Language, Galician Language, English Language, Knowledge of the Environment, and Mathematics. Average achievement was calculated with the mean grades in these five areas.
Data of the target variables was collected during regular school hours, by research assistants, after obtaining the consent of the school administration and of the teachers and students. Prior to the application of the questionnaires, which took place in a single session, the participants were informed about the goals of the project, and assured that data was confidential and used for research purposes only.
The model was fit with AMOS 18 (Arbuckle, 2009). The data were previously analyzed and individual cases presenting a significant number of missing values were eliminated (2.1%), whereas the rest of the missing values were replaced by the mean. Taking into account the analysis of the characteristics of the variables (e.g., skewness and kurtosis in Table 1), we used the maximum likelihood method to fit the model and estimate the values of the parameters.
Means, standard deviations, skewness, kurtosis, and correlation matrix of the target variables.
A series of goodness-of-fit statistics were used to analyze our model. Beyond chi-square (χ2) and its associated probability (p), the information provided by the goodness-of-fit index (GFI) and the adjusted goodness-of-fit index (AGFI; Jöreskog and Sörbom, 1983); the comparative fit index (CFI) (Bentler, 1990); and the root mean square error of approximation (RMSEA; Browne and Cudeck, 1993) was used. According to these authors, the model fits well when GFI and AGFI > 0.90, CFI > 0.95, and RMSEA ≤ 0.05.
The relations between the variables included in the model as well as the descriptive statistics are shown in Table 1. All the variables were significantly and positively related, except for the time spent on homework, which was only related to the amount of homework done. According to the value of the means of these variables, students in the last years of elementary school: (a) reported a high level of motivation to learn and mastery; (b) used preferentially a deep approach to homework; (c) did the homework assigned by the teachers most of the times; (d) usually spent about an hour a day on homework; (e) reported to manage their study time effectively; and (f) showed a medium-high level of academic achievement.
Evaluation and re-specification of the initial model
The data obtained indicated that the initial model (see Figure 1) presented a poor fit to the empirical data: χ2 = 155.80, df = 8, p < 0.001, GFI = 0.917, AGFI = 0.783, TLI = 0.534, CFI = 0.751, RMSEA = 0.186, 90% CI (0.161, 0.212), p < 0.001. Analysis of the modification indexes revealed the need to include three direct effects initially considered as null, and to eliminate a finally null effect (included in the initial model as significant). The strategy adopted to modify the initial model involved including and estimating the model each time a new effect was included. The final model comprised three effects (academic goals on homework time management, on amount of homework done, and on academic achievement) and the elimination of the initially established effect of the approach to studying on the time spent doing homework. The inclusion or elimination of the effects in the model was determined accounting for their statistical and theoretical significance. The final model resulting from these modifications is shown in Figure 2, with an adequate fit to the empirical data: χ2 = 12.03, df = 6, p = 0.061, GFI = 0.993, AGFI = 0.974, TLI = 0.975, CFI = 0.990, RMSEA = 0.043, 90% CI (0.000, 0.079), p = 0.567.
The results of the fit of the hypothesized model (standardized outcomes): Relations in dashed lines were found to be statistically significant, but this was not established in the initial model.
Assessment of the relationships on the final model
Table 2 presents the data obtained for the relationships considered in the final model (see also Figure 2).
Fit of the hypothesized model (standardized outcomes): final model of student engagement in homework.
The data from Table 2 and Figure 2 indicates that the majority of the relationships between the variables are consistent with the hypotheses. First, we found a statistically significant association between the learning goals (i.e., competence and control), the approach to homework (b = 0.50, p < 0.001), two of the variables associated with engagement in homework (the amount of homework done [b = 0.27, p < 0.001], homework time management [b = 0.30, p < 0.001]), and academic achievement (b = 0.34, p < 0.001). These results indicate that the more oriented students are toward learning goals (i.e., competence and control), the deeper the approach to homework, the more homework is completed, the better the homework time management, and the higher the academic achievement.
Second, a statistically significant association between the deep approach and homework time management (b = 0.30, p < 0.001) and the amount of homework done (b = 0.09, p < 0.05) was found. These results reflect that the deeper the students' approach to homework, the better the management of the time spent on homework, and the more the homework done. Third, there was a statistically significant association between homework time management, time spent on homework, and the amount of homework done (b = 0.23, p < 0.001, and b = 0.10, p < 0.01, respectively). These results confirm, as expected, that the more time students spent doing homework and the better students manage their homework time, the more homework they will do. Four, we found a statistically significant relation between the amount of homework done and academic achievement (b = 0.20, p < 0.001). This indicates that the more homework students complete the better their academic achievement.
In summary, our findings indicate that: (a) academic achievement is positively associated with the amount of homework completed; (b) the amount of homework done is related to homework time management; (c) homework time management is associated with how homework is done (approach to homework); and (d) consistent with the behavior of the variables in the model (except for the time spent on homework), how homework is done (i.e., approach to homework) is explained to a great extent (see total effects in Table 3) by the student's type of academic motivation.
Standardized direct, indirect, and total effects for the final model.
Finally, taking into account both the direct effects (represented in Figure 2) and the indirect ones (see Table 3), the model explained between 20 and 30% of the variance of the dependent variables (except for the time spent on homework, which is not explained at all): approach to homework (24.7%), time management (26.9%), amount of homework done (24.4%), and academic achievement (21.6%).
Consistent with prior research (e.g., Cooper et al., 2001), our findings showed that students' academic achievement in the last years of elementary education is closely related to the amount of homework done. In addition, the present study also confirms the importance of students' effort and commitment to doing homework (Trautwein et al., 2006a,b), showing that academic achievement is also related with students' desire and interest to learn and improve their skills. Therefore, when teachers assign homework, it is essential to attend to students' typical approach to learning, which is mediated by the motivational profile and by the way students solve the tasks proposed (Hong et al., 2004). The results of this investigation suggest that the adoption of learning goals leads to important educational benefits (Meece et al., 2006), among which is doing homework.
Importantly, our study shows that the amount of homework done is associated not only with the time spent, but also with the time management. Time spent on homework should not be considered an absolute indicator of the amount of homework done, because students' cognitive skills, motivation, and prior knowledge may significantly affect the time needed to complete the homework assignment (Regueiro et al., 2015). For students, managing homework time is a challenge (Corno, 2000; Xu, 2008), but doing it correctly may have a positive influence on their academic success (Claessens et al., 2007), on homework completion (Xu, 2005), and on school achievement (Eilam, 2001).
Despite, that previous studies reported a positive relationship between the time spent on homework and academic achievement (Cooper et al., 2006), the present research shows that time spent on homework is not a relevant predictor of academic achievement. Other studies have also obtained similar results (Trautwein et al., 2009; Núñez et al., 2015a), indicating that time spent on homework is negatively associated to academic achievement, perhaps because spending a lot of time on homework may indicate an inefficient working style and lack of motivation (Núñez et al., 2015a). Besides, our data indicates that spending more time on homework is positively associated to the amount of homework done.
Although, some studies have found that students who spend more time on homework also tend to report greater commitment to school work (Galloway et al., 2013), our findings indicated that spending more time doing homework was not related to a deeper engagement on the task. A possible explanation may be that using a deep approach to school tasks subsumes engaging in homework with the aim of practicing but also to further extend the content learned in class. This approach does not depends on the time spent doing homework, rather on the students' motives for doing homework.
Another important contribution of this study concerns learning-oriented goals—usually associated with positive outcomes in motivational, cognitive, and achievement variables (Pajares et al., 2000). Results indicate that the motivation to increase competence and learning is also related to approaching homework deeply and to manage homework efficiently. Consistent with previous findings (Xu, 2005), these results provide additional empirical support to time management goals (Pintrich, 2004).
There is a robust relationship between learning-oriented goals and a deep approach, and between a deep approach and the amount of homework done. All this indicates that these results are in line with prior research, meaning that the adoption of a deep approach to learning is related with high quality academic achievement (Lindblom-Ylänne and Lonka, 1999; Rosário et al., 2013b).
Educational implications and study limitations
One of the major limitations of this study lies in the type of research design used. We used a cross-sectional design to examine the effects among the variables within a path analysis model. However, to establish a cause-effect relationship a temporal sequence between two variables is needed a requirement that can only be met with longitudinal designs. Future studies should consider address this limitation.
Despite the above limitation, our results can be considered relevant and show important educational implications. It is essential for teachers and school administrators to be sensitized about the effects of teachers' homework follow-up practices on students' homework engagement (Rosário et al., 2015), and of these variables in students' school engagement and academic success. Likewise, research on students' learning should be undertaken from the perspective of the learners to understand how students use their knowledge and skills to do homework and to solve problems posed therein. On the other hand, research should examine in-depth the use of learning strategies during homework, as well as how students' motivations at an early age may foster homework completion and increase the quality of school outcomes. For this last purpose, teachers should pay attention not only to the acquisition of curricular content but also to the development of the appropriate thinking skills and self-regulated learning strategies (Rosário et al., 2010b; Núñez et al., 2013). Finally, the amount of homework done and its positive relationship with academic achievement should be considered as a final outcome of a process rooted on a comprehensive and meaningful learning. Students motivated to learn are likely to approach homework deeply and manage homework time efficaciously. As a result, they tend to do more homework and outperform. In sum, is doing homework a good way to acquire competence, improve skills, and outperform? Our data suggest a positive answer.
AV and BR Collect data, data analysis, writing the paper. JN and PR data analysis, writing the paper. SR and IP writing the paper.
Conflict of interest statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This work was developed through the funding of the research project EDU2013-44062-P, of the State Plan of Scientific and Technical Research and Innovation 2013-2016 (MINECO) and to the financing received by one of the authors in the FPU program of the Ministry of Education, Culture, and Sport.
- Appleton J. J., Christenson S. L., Furlong M. J. (2008). Student engagement with school: critical conceptual and methodological issues of the construct. Psychol. Sch.45, 369–386. 10.1002/pits.20303 [Cross Ref]
- Appleton J. J., Christenson S. L., Kim D., Reschly A. L. (2006). Measuring cognitive and psychological engagement: validation of the student engagement instrument. J. Sch. Psychol.44, 427–445. 10.1016/j.jsp.2006.04.002 [Cross Ref]
- Arbuckle J. L. (2009). Amos 18.0 User's Guide. Crawfordville, FL: Amos Development Corporation.
- Bentler P. M. (1990). Comparative fit indexes in structural models. Psychol. Bull.107, 238–246. 10.1037/0033-2909.107.2.238 [PubMed][Cross Ref]
- Biggs J., Kember D., Leung D. Y. (2001). The revised two-actor study process questionnaire: R-SPQ-2F. Br. J. Educ. Psychol.71, 133–149. 10.1348/000709901158433 [PubMed][Cross Ref]
- Boekaerts M. (1999). Self-regulated learning: where are today. Int. J. Educ. Res.31, 445–458. 10.1016/S0883-0355(99)00014-2 [Cross Ref]
- Browne M. W., Cudeck R. (1993). Alternative ways of assessing model fit, in Testing Structural Equation Models, eds Bollen K., Long J., editors. (Newbury Park, CA: Sage; ), 136–162.
- Cano F., García A., Justicia F., García-Berbén A. B. (2014). Enfoques de aprendizaje y comprensión lectora: el papel de las preguntas de los estudiantes y del conocimiento previo [Approaches to learning and reading comprehension: the role of students' questions and of prior knowledge]. Rev. Psicodidáctica19, 247–265. 10.1387/RevPsicodidact.10186 [Cross Ref]
- Claessens B. J. C., van Eerde W., Rutte C. G., Roe R. A. (2007). A review of the time management literature. Pers. Rev.36, 255–276. 10.1108/00483480710726136 [Cross Ref]
- Cooper H., Jackson K., Nye B., Lindsay J. J. (2001). A model of homework's influence on the performance evaluations of elementary school students. J. Exp. Educ.69, 181–200. 10.1080/00220970109600655 [Cross Ref]
- Cooper H., Robinson J. C., Patall E. A. (2006). Does homework improve academic achievement? A synthesis of research, 1987–2003. Rev. Educ. Res.76, 1–62. 10.3102/00346543076001001 [Cross Ref]
- Corno L. (2000). Looking at homework differently. Element. Sch. J.100, 529–548. 10.1086/499654 [Cross Ref]
- Deci E. L., Ryan R. M. (2002). Handbook of Self-Determination Research. New York, NY: University of Rochester Press.
- Dettmers S., Trautwein U., Lüdtke O. (2009). The relationship between homework time and achievement is not universal: evidence from multilevel analyses in 40 countries. Sch. Eff. Sch. Improv.20, 375–405. 10.1080/09243450902904601 [Cross Ref]
- Eccles J., Wang M. T. (2012). Part I Commentary: so what is student engagement anyway?,in Handbook of Research on Student Engagement, eds Christenson S. L., Reschly A. L., Wylie C., editors. (New York, NY: Springer; ), 133–145.
- Eccles (Parsons) J., Adler T. F., Futterman R., Goff S. B., Kaczala C. M., Meece J. L., et al. (1983). Expectancies, values, and academic choice: origins and changes, in Achievement and Achievement Motivation, ed Spence J., editor. (San Francisco, CA: Freeman; ), 75–146.
- Eilam B. (2001). Primary strategies for promoting homework performance. Am. Educ. Res. J.38, 691–725. 10.3102/00028312038003691 [Cross Ref]
- Elliot A. J. (1999). Approach and avoidance motivation and achievement goals. Educ. Psychol.34, 169–189. 10.1207/s15326985ep3403_3 [Cross Ref]
- Elliot A. J., Church M. A. (1997). A hierarchical model of approach and avoidance achievement motivation. J. Pers. Soc. Psychol.72, 218–232. 10.1037/0022-35188.8.131.52 [Cross Ref]
- Elliot A. J., McGregor H. A., Gable S. (1999). Achievement goals, study strategies, and exam performance: a mediational analysis. J. Educ. Psychol.91, 549–563. 10.1037/0022-06184.108.40.2069 [Cross Ref]
- Entwistle N. J. (1991). Approaches to learning and perceptions of the learning environment. High. Educ.22, 201–204. 10.1007/BF00132287 [Cross Ref]
- Entwistle N. J. (2009). Teaching for Understanding at University: Deep Approaches and Distinctive Ways of Thinking. Basingstoke: Palgrave Macmillan.
- Fernández-Alonso R., Suárez-Álvarez J., Muñiz J. (2014). Tareas escolares en el hogar y rendimiento en matemáticas: una aproximación multinivel con estudiantes de Enseñanza Primaria [Homework and academic performance in mathematics: a multilevel approach with Primary school students]. Rev. Psicol. Educ.9, 15–29.
- Fredricks J. A., Blumenfeld P. C., Paris A. (2004). School engagement: potential of the concept, state of the evidence. Rev. Educ. Res.74, 59–109. 10.3102/00346543074001059 [Cross Ref]
- Galloway M., Conner J., Pope D. (2013). Nonacademic effects of homework in privileged, high-performing high schools. J. Exp. Educ.81, 490–510. 10.1080/00220973.2012.745469 [Cross Ref]
- Gaudreau P. (2012). Goal self-concordance moderates the relationship between achievement goals and indicators of academic adjustment. Learn. Individ. Differ.22, 827–832. 10.1016/j.lindif.2012.06.006 [Cross Ref]
- Hong E., Milgram R. M., Rowell L. L. (2004). Homework motivation and preference: a learner-centered homework approach. Theory Pract.43, 197–203. 10.1207/s15430421tip4303_5 [Cross Ref]
- Inglés C. J., Martínez-González A. E., García-Fernández J. M. (2013). Conducta prosocial y estrategias de aprendizaje en una muestra de estudiantes españoles de Educación Secundaria Obligatoria [Prosocial behavior and learning strategies in a sample of Spanish students of Compulsory Secondary Education]. Eur. J. Educ. Psychol.6, 33–53. 10.1989/ejep.v6i1.101 [Cross Ref]
- Jöreskog K. G., Sörbom D. (1983). LISREL - 6 User's Reference Guide. Mooresville, IN: Scientifi c Software.
- Kohn A. (2006). Abusing research: the study of homework and other examples. Phi Delta Kappan88, 9–22. 10.1177/003172170608800105 [Cross Ref]
- Lindblom-Ylänne S., Lonka K. (1999). Individual ways of interacting with the learning environment - are they related to study success?Learn. Instruct.9, 1–18. 10.1016/S0959-4752(98)00025-5 [Cross Ref]
- Marks H. (2000). Student engagement in instructional activity: patterns in the elementary, middle, and high school years. Am. Educ. Res. J.37, 153–184. 10.3102/00028312037001153 [Cross Ref]
- Martin A. J. (2012). Motivation and engagement: conceptual, operational, and empirical clarity, in Handbook of Research on Student Engagement, eds Christenson S. L., Reschly A. L., Wylie C., editors. (New York, NY: Springer; ), 303–311.
- Marton F., Säljö R. (1976a). On qualitative differences in learning. I: outcome and process. Br. J. Educ. Psychol.46, 4–11. 10.1111/j.2044-8279.1976.tb02980.x [Cross Ref]
- Marton F., Säljö R. (1976b). On qualitative differences in learning. II: outcome as a function of the learner's conception of the task. Br. J. Educ. Psychol.46, 115–127. 10.1111/j.2044-8279.1976.tb02304.x [Cross Ref]
- Meece J. L., Anderman E. M., Anderman L. H. (2006). Classroom goal structure, student motivation, and academic achievement. Annu. Rev. Psychol.57, 487–503. 10.1146/annurev.psych.56.091103.070258 [PubMed][Cross Ref]
- Middleton M., Midgley C. (1997). Avoiding the demonstration of lack of ability: an unexplored aspect of goal theory. J. Educ. Psychol.89, 710–718. 10.1037/0022-06220.127.116.110 [Cross Ref]
- Muhlenbruck L., Cooper H., Nye B., Lindsay J. J. (2000). Homework and achievement: explaining the different strengths of relation at the elementary and secondary school levels. Soc. Psychol. Educ.3, 295–317. 10.1023/A:1009680513901 [Cross Ref]
- Ng C. H. (2008). Multiple-goal learners and their differential patterns of learning. Educ. Psychol.28, 439–456. 10.1080/01443410701739470 [Cross Ref]
- Núñez J. C., González-Pienda J. A., González-Pumariega S., García M., Roces C. (1997). Cuestionario Para la Evaluación de Metas Académicas [Academic Goals Assessment Questionnaire]. Department of Psychology, University of Oviedo.
- Núñez J. C., Suárez N., Cerezo R., González-Pienda J. A., Rosário P., Mourão R., et al. (2015a). Homework and academic achievement across Spanish Compulsory Education. Educ. Psychol.35, 726–746. 10.1080/01443410.2013.817537 [Cross Ref]
- Núñez J. C., Suárez N., Rosário P., Vallejo G., Cerezo R., Valle A. (2015b). Teachers' feedback on homework, homework-related behaviors and academic achievement. J. Educ. Res.108, 204–216. 10.1080/00220671.2013.878298 [Cross Ref]
- Núñez J. C., Suárez N., Rosário P., Vallejo G., Valle A., Epstein J. L. (2015c). Relationships between parental involvement in homework, student homework behaviors, and academic achievement: differences among elementary, junior high, and high school students. Metacogn. Learn.10, 375–406. 10.1007/s11409-015-9135-5 [Cross Ref]
- Núñez J., Rosário P., Vallejo G., González-Pienda J. (2013). A longitudinal assessment of the effectiveness of a school-based mentoring program in middle school. Contemp. Educ. Psychol.38, 11–21. 10.1016/j.cedpsych.2012.10.002 [Cross Ref]
- Pajares F., Britner S. L., Valiante G. (2000). Relation between achievement goals and self-beliefs of middle school students in writing and science. Contemp. Educ. Psychol.25, 406–422. 10.1006/ceps.1999.1027 [PubMed][Cross Ref]
- Pintrich P. R. (2004). A conceptual framework for assessing motivation and self-regulated learning in college students. Educ. Psychol. Rev.16, 385–407. 10.1007/s10648-004-0006-x [Cross Ref]
- Pintrich P. R., De Groot E. V. (1990). Motivational and self-regulated learning components of classroom performance. J. Educ. Psychol.82, 33–40. 10.1037/0022-0618.104.22.168 [Cross Ref]
- Reeve J., Jang H., Carrell D., Jeon S., Barch J. (2004). Enhancing students' engagement by increasing teachers' autonomy support. Motiv. Emot.28, 147–169. 10.1023/B:MOEM.0000032312.95499.6f [Cross Ref]
- Regueiro B., Suárez N., Valle A., Núñez J. C., Rosário P. (2015). La motivación e implicación en los deberes escolares a lo largo de la escolaridad obligatoria [Homework motivation and engagement throughout compulsory education]. Rev. Psicodidáctica20, 47–63. 10.1387/RevPsicodidact.12641 [Cross Ref]
- Rodríguez S., Cabanach R. G., Piñeiro I., Valle A., Núñez J. C., González-Pienda J. A. (2001). Metas de aproximación, metas de evitación y múltiples metas académicas [Approach goals, avoidance goals and multiple academic goals]. Psicothema13, 546–550.
- Rosário P., González-Pienda J. A., Pinto R., Ferreira P., Lourenço A., Paiva O. (2010a). Efficacy of the program “Testas's (mis)adventures” to promote the deep approach to learning. Psicothema22, 828–834. [PubMed]
- Rosário P., Mourão R., Baldaque M., Nunes T., Núñez J. C., González-Pienda J. A., et al. (2009). Homework, self-regulation of learning and math performance. Rev. Psicodidáctica14, 179–192.
- Rosário P., Núñez J. A., Ferrando J. P., Paiva O., Lourenço A., Cerezo R., et al. (2013a). The relationship between approaches to teaching and approaches to studying: a two-level structural equation model for biology achievement in high school. Metacogn. Learn.8, 47–77. 10.1007/s11409-013-9095-6 [Cross Ref]
- Rosário P., Núñez J. C., González-Pienda J. A., Almeida L., Soares S., Rúbio M. (2005). Academic learning from the perspective of Model 3P of J. Biggs. Psicothema17, 20–30.
- Rosário P., Núñez J. C., González-Pienda J. A., Valle A., Trigo L., Guimarães C. (2010b). Enhancing self-regulation and approaches to learning in first-year college students: a narrative-based program assessed in the Iberian Peninsula. Eur. J. Psychol. Educ.25, 411–428. 10.1007/s10212-010-0020-y [Cross Ref]
- Rosário P., Núñez J. C., Vallejo G., Cunha J., Azevedo R., Pereira R., et al. (2016). Promoting Gypsy children school engagement: a story-tool project to enhance self-regulated learning. Contemp. Educ. Psychol. [Epub ahead of print]. 10.1016/j.cedpsych.2015.11.005. [Cross Ref]
- Rosário P., Núñez J. C., Vallejo G., Cunha J., Nunes T., Suárez N., et al. . (2015). The effects of teachers' homework follow-up practices on students' EFL performance: a randomized-group design. Front. Psychol.6:1528. 10.3389/fpsyg.2015.01528 [PMC free article][PubMed][Cross Ref]
- Rosário P., Núñez J., Valle A., González-Pienda J., Lourenço A. (2013b). Grade level, study time, and grade retention and their effects on motivation, self-regulated learning strategies, and mathematics achievement: a structural equation model. Eur. J. Psychol. Educ.28, 1311–1331. 10.1007/s10212-012-0167-9 [Cross Ref]
- Skaalvik E. (1997). Self- enhancing and self-defeating ego orientation: relations with task and avoidance orientation, achievement, self- perceptions, and anxiety. J. Educ. Psychol.89, 71–81. 10.1037/0022-0622.214.171.124 [Cross Ref]
- Skinner E. A., Pitzer J. R. (2012). Developmental dynamics of student engagement, coping, and everyday resilience, in Handbook of Research on Student Engagement, eds Christenson S. L., Reschly A. L., Wylie C., editors. (New York, NY: Springer; ), 21–44.
- Struyven K., Dochy F., Janssens S., Gielen S. (2006). On the dynamics of students' approaches to learning: the effects of the teaching/learning environment. Learn. Instr.16, 279–294. 10.1016/j.learninstruc.2006.07.001