The academic support can be understood as the emotional and instrumental support given by significant others to the student in the academic context (Alfaro et al., 2006; Chen, 2005). Generally, when we speak about academic support we refer to the academic support perceived by students, regardless of its possible incongruence with the perception by teachers, parents or peers (Choe, 2020). The influence that support from the teachers, families, or peers has on adolescents’ academic lives has been the object of study in many recent research studies (Gutiérrez et al., 2017), with several of these showing their roles as protective factors against the development of anxiety or depression (Arora et al., 2017; Boudreault-Bouchard et al., 2013; Leung et al., 2010; Smokowski et al., 2015, Rueger et al., 2016). Specifically, Lei et al. (2018) carried out a meta-analysis in order to explore the relationship between teacher support and students’ emotions. After evaluating 65 studies they concluded that students with more support from teachers showed more positive emotions.

In addition to the importance of support (from various sources) to the personal development of adolescents, there is also evidence for its relevance in academic domains. Specifically, evidence has been uncovered that suggests a relationship between academic support from parents, teachers and/or peers and academic achievement (Jelas et al., 2016; Shen et al., 2014; Lam et al., 2012; Wentzel et al., 2016; Wang & Eccles, 2012), academic motivation (Alfaro et al., 2006; Alfaro & Umaña-Taylor, 2015; Horanicova et al., 2020; Plunkett & Bámaca-Gómez, 2003; Sands & Plunkett, 2005), and school engagement or involvement (Clark et al., 2020; Estell & Perdue, 2013; Gutiérrez et al., 2017; Mikami et al., 2017; Ramos-Díaz et al., 2016, Virtanen et al., 2014). When teachers were respectful and took care of their students, these young people seemed to be more committed to studying the subject in question and were more polite to the rest of the class (Chiu & Chow, 2011; Longobardi et al., 2016). Furthermore, there is also evidence for the opposite effect: when teachers are disrespectful, students were less cooperative (Miller et al., 2000).

Based on traditional educational theories as the attribution theory, expectancy-value theory, goal theory, self-determination theory, self-efficacy or self-worth motivation theory, positive interpersonal relationships could affect academic motivation by directly impact its base: students’ beliefs and emotions (for an extensive review see Martin and Dowson, 2009). Through these relationships, students internalize the beliefs valued by others and increase their sense of belonging, which translates into greater motivation and self-esteem. Recent studies have empirically deepened the study of the effect of academic support and academic motivation. Horanicova et al. (2020) evidenced that the support from teachers and peers improves student’s satisfaction with schools and their education’s relevance perception. These results would be consistent across genders (Horanicova et al., 2020). The importance of academic support would be especially relevant on students with low family socioeconomic status, where the academic support acts as a buffer of its negative impact on academic engagement and attitudes towards education (Horanicova et al., 2022). Although previous literature has shown the relevance of support from home and school, peer support seems to be less important than other sources of support (Saleh et al., 2019, Sethi & Scales, 2020). This relationship with motivation is especially relevant given that through it, academic support has an indirect impact on academic performance (Sethi & Scales, 2020).

Bronfenbrenner’s (1989) ecological model is frequently used in the literature to help understand the influence of contextual variables such as support from teachers, families, or the student peer group in adolescents’ development. This model highlights the importance of understanding the development of people as a process of constant interaction between individuals and their environments. Specifically, Bronfenbrener (1989) emphasised the importance of ‘significant others’, understood as individuals in the person’s immediate environment who directly influence them. Regarding adolescents, these significant others usually correspond to their teachers, parents, and peers (Alfaro, et al., 2006) as socialisation agents; the resources they provide to students to facilitate their academic achievement can be defined as ‘academic support’ (Alfaro et al., 2006). This academic support may be bestowed in different ways such as emotional support (understanding a student’s feelings and encouraging them) or instrumental support (helping them with homework and to understand the topics they are studying) (Chen, 2005). These factors, emotional and instrumental support, are included in the framework proposed specifically for the teacher-student interactions by Hamre and Pianta (2007), and tested by Hamre et al. (2013), which also includes classroom organization as a relevant factor in the students learning promotion.

There are several different approaches to academic support and these have resulted in different scales and studies that, despite measuring the same thing, do so with differentiating nuances (Malecki & Elliott, 1999). Additionally, most research has focused on measuring the influence of the support from one of these groups (Cattley, 2004), rather than taking a multidimensional approach to the concept, jointly considering different socialisation agents as sources of academic support. Specifically, a large volume of research has explored the relationship between family and parental support and the academic development of students, but less attention has focussed on their peers (Furrer & Skinner, 2003). However, there is broad consensus that social support has a multidimensional nature, and that its various dimensions probably relate in different ways to particular types of outcomes (Sarason & Sarason 2009). Thus, continued investigation to expand our understanding of this area and to improve the measurement instruments available will enable rigorous research in the field of educational psychology to understand the effect of academic support when teacher, parental and peer support are considered.

Our literature review highlighted the availability of several interesting instruments for studying academic support. For example, the Student Social Support Scale (SSSS; Nolten, 1994; Malecki & Elliott, 1999), based on Tardy’s (1985) social support model, measures emotional, informational, instrumental, and appraisal support from teachers, parents, peers, and close friends. This scale contains a total of 60 items and its language is aimed at primary school students rather than adolescents. To overcome these limitations, Malecki et al. (1999) developed the Scale of Social Support for Children and Adolescents (CASSS). Despite having similarities with the SSSS, the CSASSS is comprised of 40 items and includes a version for children and adolescents. With satisfactory results from their validation (Malecki & Demaray, 2002) studies, both these scales and adapted versions of them continue to be widely used (Ciarrochi et al., 2017; Maiuolo et al.; 2019; Archer et al., 2019), but they are also both considered to be too long for administration alongside other instruments. Additionally, they measure social support in the academic field rather than academic support as defined above, even though both these factors are closely related.

Thompson and Mazer (2009) and Mazer and Thompson (2011) subsequently proposed an instrument focused on academic support, the Student Academic Support Scale (SASS). This scale was designed with university students in mind and exclusively measures peer support. These authors developed another questionnaire for parental academic support which was oriented towards individuals not yet at an undergraduate level, the Parental Academic Support Scale (PASS; Thompson & Mazer, 2012; Mazer & Thompson, 2016). The PASS understands parental academic support as the frequency of communication between parents and teachers about academic performance, classroom behaviour, preparation, hostile interactions, and student health. However, once again, these scales are specific of peers or parent support rather than multidimensional, and in the latter case, does not focus on the support perceived by students.

Another contribution of special relevance in this arena is the scale developed by Sands and Plunkett (2005) to measure the academic support derived from the mothers, fathers, teachers, and friends of Latino adolescents in the United States. As one of its differentiating characteristics, the Significant Other Academic Support Scale registers students’ perceptions of their mothers and fathers separately. Gutiérrez et al. (2017) generated a dimension termed ‘family support’ by combining the scores obtained from the separate father and mother support factors on this scale. In their research, these authors then measured family support using this dimension, as well as teacher and peer support by adapting two subscales from Lam et al. (2012).

In summary, although academic support has been evidenced as a relevant construct in the academic and personal development of students, a brief and integrated instrument that considers different sources of support (teacher, family and peers) is still required to measure it. Following the adaptations suggested by Gutiérrez et al. (2017), here we propose a scale of perceived academic support, the Perceived Academic Support Questionnaire (PASQ): that includes the dimensions of family support (based on Sands and Plunkett (2005)), teacher support, and peer support (based on Lam et al., 2012).

Purpose of this study

This study aims to validate a support instrument based on the aforementioned scales in a sample of high school students in Dominican Republic. In the context of the Dominican Republic, and in spite of having implemented advances in educational improvement and investment (Acción Empresarial para la Educación, EDUCA, 2015; Iniciativa Dominicana por una Educación de Calidad, IDEC, 2014), the country continues to show low academic achievement rates (Organización para la Cooperación y el Desarrollo Económico, OECD, 2016). Having an easy-to-use tool to detect the levels of social support (from parents, teachers and peers) would be a step forward in the detection of possible obstacles to the improvement of academic performance.

Therefore, the specific aims of the study includes: (1) to test the three-factor structure validity of the PASQ in Dominican Republic students; (2) to test the reliability of the PASQ; and (3) to test the external validity by proving its predictive power on academic motivation, doing so we expect to find empirical support for the theoretically expected relationships in order to obtain evidences for the interpretation of the results.

Method

Participant

Our sample comprised 1,712 secondary students from educational districts 04–03 and 11–01 of the Dominican Republic. All the regions of Dominican Republic were contacted and finally the study was carried out where the authorities showed interest and availability to participate. Within the available regions, districts were chosen for the presence of significant challenges in terms of their indicators of academic success. The inclusion criteria for participants were: a) being a student from an educational center in districts 04-03 or 11-01; b) being a high school student; and c) consenting to participate. The aforementioned districts showed a population of 3387 students, considering a level of confidence of 99% and 3% margin of error (with p = q = 0.5), 1712 students were sampled. The lack of response from them was negligible (less than 1%).

Of the total sample, 902 were female (52.69%) and 809 were male (47.25%), one student did not declare gender. All the participants were aged between 12 and 20 years (mean = 14.73 ± 1.2 years). A total of 1,278 students belonged to public institutions (74.65%), while 268 went to private institutions (15.65%) and 166 to semi-official educational institutions (9.70%). Of all the students, 404 were from rural areas (23.6%) and 1,308 from urban areas (76.4%). The most common family types were nuclear (52.37%) and single-parent families (32.63%), followed by extended families (9.20%) and other situations (5.80%).

Procedure

The survey procedure meets the ethical standards of the American Psychological Association (APA) and it received the ethics approval of the Ministry of Education of the Dominican Republic and the directors of the education institutions. The researchers contacted with the educational authorities to communicate the goals and procedure of the study. The schools informed teachers, families and students about the aim of the study. The schools were randomly selected among all the available ones and then invited to voluntarily participate. They did not get any reward for their participation.

The survey team was formed by three district educational technicians and two school psychologists who were in charge of distributing, supervising the implementation, and collecting the questionnaires completed in the schools. The questionnaires were administered at the schools during the first teaching hour. The students needed around 45 minutes to fulfil the self-administered questionnaire. All but 3 students completed the survey properly, whose responses were removed and replaced with the responses of 3 new students.

Instruments

We measured academic support from teachers, families, and peers using the PASQ (Cuestionario de Apoyo Académico Percibido, referred to as the CAAP in Spanish) which was originally developed in Spanish and was adapted as suggested by Gutiérrez et al. (2017). The resulting PASQ scale has a total of 12 items belonging to 3 dimensions: family, teacher, and peer support. The response scale ranges from 1 (strongly disagree) to 5 (strongly agree). Item contents are presented in the Table 1. The Spanish version of the scale is available under request.

Table 1 Descriptive statistics and the homogeneity of the items
Full size table

The family support dimension is based on the parental support dimensions from the Significant Other Academic Support Scale by Sands and Plunkett (2005), while the teacher and peer support dimensions were adapted from two of the subscales published by Lam et al. (2012), which had itself adapted items from the Caring Adult Relationships in School Scale and Caring Peer Relationships in School Scale, both from the California Healthy Kids Survey (WestEd, 2000). Gutierrez et al. (2017) reported good psychometric properties for the family support dimension in an Angolan sample: its reliability was α = 0.73 and the one-factor confirmatory factor analysis (CFA) satisfactorily fitted the data (χ2(9) = 34.56, p < 0.001, CFI = 0.995, RMSEA = 0.037). Additionally, they tested the psychometric properties of the parents and peer support scale with good results: its reliability was 0.71 and 0.70 and the two-factor structure satisfactorily fitted the data (χ2(8) = 32.74, p < 0.001, CFI = 0.992, RMSEA = 0.039).

Academic motivation was measured using the Adolescent Academic Motivation Scale used in Plunkett and Bámaca-Gómez (2003). This questionnaire contains a total of 5 Likert-type items that evaluate a single dimension (e. g. “In general, I like school”). The response scale ranges from 1 (strongly disagree) to 5 (strongly agree). In this sample, its reliability was α = 0.71 and the one-factor CFA produced a satisfactory fit: χ2(5) = 57.388 (p < 0.001), CFI = 0.975, RMSEA = 0.078, 90% CI [0.061, 0.095], SRMR = 0.024.

Statistical analysis

The descriptive statistics and reliability for the instrument were calculated using SPSS software (version 26, IBM Corp., Armonk, NY.). The descriptive statistics included the item means, standard deviations, inter-item correlations, and homogeneity; for the reliability analysis we estimated the Cronbach alpha for all the dimensions included in the scale. This reliability index is considered acceptable when it scores above 0.70, with values exceeding 0.80 being preferable (Cicchetti, 1994; Clark & Watson, 1995). Although Cronbach’s alpha is the most widely used reliability index, it has been criticised because, among other reasons, it can underestimate the true reliability (Raykov, 2004). The composite reliability index (CRI) has been shown to better estimate reliability than the Cronbach alpha (Raykov et al., 2010) and so, we decided to calculate and consider both these reliability indices.

Factorial validity was assessed via CFA. To check the model fit, we used the chi-square test, CFI, RMSEA, and SRMR. The model fit was considered acceptable if the CFI exceeded 0.90 (with 0.95 being preferable), and RMSEA and SRMR values were below 0.08 (Hu & Bentler, 1999). These analyses were carried out with Mplus software (version 8.4, Muthén & Muthén, 1998–2017) using WLSMV (Weighted Least Square Mean and Variance corrected) considering the ordinal nature of the data and the non-multivariate normality. External validity was tested relating academic support and academic motivation given that these variables have been consistently associated in previous research (Jelas et al., 2016; Shen et al., 2014; Lam et al., 2012; Wentzel et al., 2016; Wang & Eccles, 2012). We constructed a structural equation model in which the three dimensions of academic support that predicted academic motivation were tested, using the same CFI, RMSEA, and SRMR values to evaluate the overall model fit. In addition, we also examined the relationships between these variables.

Results

Factorial validity

Although previous literature recommends a three-factor model, a one-factor structure CFA was tested as a baseline model. The one-factor structure does not show a good fit to the data: χ2(54) = 2832.779 (p < 0.001), CFI = 0.794, RMSEA = 0.173, 90% CI [0.168, 0.179], and SRMR = 0.105. Consequently, a three-factor structure CFA was tested and fit as good: χ2(51) = 212.566 (p < 0.001), CFI = 0.988, RMSEA = 0.043, 90% CI [0.037, 0.049], and SRMR = 0.023 which indicated that the theoretical a priori factor structure adequately fit the data.

Figure 1 shows the tree-factor model, and the standardised loading estimates. Every indicator had a high and significant loading (p < 0.001) in its hypothesised factor. The standardised loadings ranged from a minimum of 0.690 to a maximum of 0.822. As hypothesised, the three factors were positively correlated, but the correlations were not sufficiently high to jeopardise their discriminant validity. Notwithstanding, the highest correlation was found between teacher and peer support.

Fig. 1
figure 1

Confirmatory factor analysis model of the Perceived Academic Support Questionnaire. Note: *p < 0.05

Full size image

Internal consistency and inter-item correlations

Table 1 includes means, standard deviations, homogeneity (corrected item overall correlation), skewness and kurtosis statistics of the items. The descriptive statistics and correlation among dimensions has been included in Table 2. Table 2 also includes the information for academic motivation. The Cronbach alpha was 0.69 for the teacher support dimension, 0.73 for peer support, and 0.87 for parental support. If we removed any of the items, the Cronbach alpha for the dimension decreased. In addition, these results were supplemented with the estimation of CRIs at 0.76 for teacher support, 0.79 for peer support, and 0.90 for parental support. Thus, in this sample, these results showed adequate reliability for all the dimensions.

Table 2 Descriptive statistics and correlations among dimensions
Full size table

External validity

To assess external validity, we tested a structural equation model with latent factors in which the latent factor of academic motivation was predicted by the three dimensions of academic support. The goodness of fit indices for these analyses were satisfactory: χ2(113) = 497.098 (p < 0.001), CFI = 0.978, RMSEA = 0.045, 90% CI [0.041, 0.049], and SRMR = 0.029. The analytical results of this model (Fig. 2) showed that the three dimensions of academic support predicted 59.1% of the variance in academic motivation (R2 = 0.591); all these relationships were statistically significant. Parental and teacher academic support positively predicted academic motivation, with structural coefficients of β = 0.607 (p < 0.001) and β = 0.369 (p < 0.001), respectively. However, peer support was negatively related to academic motivation (β = −0.145, p = 0.002).

Fig. 2
figure 2

Predictive model of academic motivation based on the three dimensions of academic support. Note: *p < 0.05

Full size image

Discussion

It has often been pointed out in the scientific literature that the academic support that students receive is an important factor in their academic and personal development. The academic support given by teachers, families, and peers positively affects student success, motivation, engagement, and well-being (Gutiérrez et al., 2017; Jelas et al., 2016; Sands & Plunkett, 2005; Tomás et al., 2020), while its absence has been related to lower educational expectations and higher levels of school truancy (Al-Alwan, 2014; Boudreau et al., 2004; Salmela-Aro & Upadyaya, 2012; Yang, 2004). Consequently, considering the importance of these factors in education, brief instruments to measure them, that integrate the dimensions of teacher, family, and peer support, are required to continue rigorous research in educational psychology. The support of various socialization agents should be taken into account in order to be able to disentangle the specific role of each one and check if any of them is more relevant than the rest in the promotion of specific educational variables. Thus, the Perceived Academic Support Questionnaire (PASQ) (in its original Spanish version, the Cuestionario de Apoyo Académico Percibido or CAAP) was developed to address this need. The PASQ incorporates the contributions of previous studies and has demonstrated adequate psychometric properties in a sample of students from the Dominican Republic.

Regarding factor validity, the CFA carried out for the a priori three-factor structure showed particularly good overall fit indices and there was evidence that the data reproduced the proposed theoretical model. Results showed positive and statistically significant correlations between the three factors or sources of academic support. Indeed, contextual variables such as support from teachers, families, or the student peer group are not independent elements, but are highly interrelated. Therefore, an instrument such as the PASQ allows to cover the three most important sources of support in adolescents’ development at a time, while also considering their interdependency.

Regarding reliability estimates, two indices were used to analyse the internal consistency of the PASQ. Firstly, Cronbach alphas produced adequate results for all the dimensions except for teacher support which fell below the cut-off criteria of 0.7 although it was remarkably close to it with a score of 0.69. Although the Cronbach alpha index has several limitations—such as assuming tau-equivalence or being a lower bound for the true reliability (Raykov, 2004)—it is still the most widely used tool for this purpose. To compensate for these flaws, we calculated the CRI for each dimension, which also exceeded the cut-off criteria and produced adequate results in all the dimensions. Reliability results were slightly better than the results obtained by Gutiérrez et al. (2017) with a sample of Angolan students.

As proof of the predictive validity of the PASQ, its relationship with academic motivation was analysed via a structural equation model in which the three dimensions of academic support predicted the latent factor of academic motivation. Significant positive correlations were obtained between parental and teacher academic support and academic motivation. These results concur with those previously reported in the literature (Alfaro et al., 2006; Isik et al., 2018; Horanicova et al., 2020; Plunkett & Bámaca-Gómez, 2003; Sands & Plunkett, 2005; Song et al. 2015), with a significant negative relationship being identified between peer support and academic motivation. However, the relationship between these latter two variables remains somewhat controversial. While many studies have found a positive relationship between peer support, engagement, and positive attitudes towards studying, it was less clear in other works (Li et al., 2010; Saleh et al., 2019; Sethi & Scales, 2020; Veentra et al., 2010; Virtanen et al., 2014). Sethi & Scales (2020) carried out a study with high school students in the United States. Their results evidenced that teacher and parental support were highly relevant to increase academic motivation and, indirectly, academic results. However, although it benefits the school climate, peer support has no statistically significant effect on academic motivation. We should bear in mind that the results of studies that consider peer support as the only source of academic support and those that include other sources of support may vary significantly, because the former do not control for the effect of parental and teacher support. The negative influence of peer support on academic motivation might be possible in a context where a general atmosphere of disruptive attitudes among peers is generated and tolerated (Mathys et al., 2013; Rubin et al., 2007).

In summary, in this current work the PASQ showed adequate psychometric properties that support its use in future research. Because of its relationship with academic motivation, this instrument will likely be useful to help detect students with inadequate support (i.e., low levels of teacher or parental support), to help prevent future difficulties in academic achievement. This is of special interest for those contexts in which students’ have poor levels of academic achievement, such as the one under study, the Dominican Republic. The Dominican Republic has repeatedly shown low levels of academic success when compared to other countries (i.e., OECD, 2019). With the assessment of academic support, we would be able to detect those adolescents who may have lower academic motivation and, therefore, problems with academic achievement. The puzzle of academic achievement is greater than a measurement scale, but every piece is welcomed to accomplish the goal.

It is important to note the limitations of this work. For example, because of the cross-sectional design of this research, we were unable to test the longitudinal invariance of the PASQ. This aspect could be examined in future studies to test the adequacy of the instrument to assess changes in academic support over time. Additionally, these current results were limited to the context of the Dominican Republic, a country that has repeatedly shown low levels of academic success relative to other countries (OECD, 2019). Indeed, only students from two districts participated, which could limit the generalizability of the results. Therefore, analysis of its performance in other cultures, languages, and education systems would be recommended in order to generalize the results to other countries. Additionally, a limitation of the scale is its self-assessment format which could be influence for social desirability and student’s response styles.