Inclusive Teaching and Learning Practices That Promote and Protect Reading and Science Literacy for Palestinian Children

Abstract

Despite adversities arising from the Palestinian experience, studies have shown that Palestinian students can demonstrate educational resilience. However, a gap in knowledge exists about inclusive teaching and learning strategies that contribute to Palestinian students’ educational resilience. This research uses mixed-effects regression modelling to examine reading and science literacy outcomes of 1279 Palestinian students in UNRWA schools in Jordan. The findings revealed two promotive factors (student online reading and teacher–student relations) and two protective factors (student enjoyment of reading and student control strategies) for reading, and two protective factors (student awareness of environmental matters and student epistemological beliefs about science) for science. Knowledge about the promotive and protective factors has important implications for reading and science educators working with Palestinian children.

1. Introduction

The United Nations Relief and Works Agency (UNRWA) was established in 1949 to aid Palestinian refugees in the occupied Palestinian territories, Jordan, Lebanon, and Syria. Throughout its history, Palestinian refugees have significantly shaped the evolution of UNRWA’s education system, advocating for curricula that reflect their socio-political realities. Despite external pressures to integrate into their host countries and abandon their Palestinian identity, Palestinian refugees have successfully campaigned for the inclusion of Palestinian history and culture in UNRWA’s educational materials, sustaining their existence as a distinct population group [1,2,3]. Over the decades, UNRWA schools grew to become a critical means for preserving Palestinian identity, while also providing a pathway towards socioeconomic mobility [4]. However, political pressure from foreign donors to depoliticize humanitarian frameworks has constrained Palestinian education, resulting in quality declines and funding cuts [3]. This reliance on depoliticized humanitarian frameworks limits education’s transformative potential, as such frameworks make no consideration for the unique adversities faced by Palestinian children [5,6].

Despite experiencing significant adversities resulting from ethnic cleansing, forced displacement, and military occupations, studies have shown that Palestinian students in UNRWA schools have demonstrated positive academic achievement and displayed educational resilience [7,8,9]. However, there is a gap in knowledge about the inclusive teaching and learning strategies that contribute to Palestinian students’ educational resilience in the presence of adversities. Accordingly, this study sought to investigate the educational resilience of Palestinian students. By utilizing mixed-effects regression modeling, this study examined the reading and science literacy outcomes of Palestinian students in UNRWA schools in Jordan, using data from PISA (Programme for International Student Assessment), an international large-scale assessment by the OECD (Organisation for Economic Co-operation and Development) that examines and compares the scholastic performance of students internationally.

Through examining the performance of Palestinian students who participated in the PISA assessments, the study intended to answer the following research questions:

• Which learning and teaching strategies function as promotive factors in reading literacy for Palestinian students?
• Which learning and teaching strategies function as protective factors in reading literacy for Palestinian students?

In doing so, the study identified several socio-ecological factors associated with positive educational outcomes [10,11], emphasizing the promotive and protective effects of these factors when interacting with risk [12,13]. Resilience factors that contributed to positive educational outcomes but whose positive effects decrease in the presence of adversity were deemed to have promotive effects, while factors that contributed to positive educational outcomes regardless of the presence of adversity were deemed to have protective effects [8,14]. Knowledge about the promotive and protective factors associated with positive reading and science literacy outcomes can have important implications on inclusive teaching and learning practices for Palestinian students.

2. Literature Review

2.1. Educational Provision for Palestinians

The Nakba (Arabic for “catastrophe”), or the ethnic cleansing of Palestinians from 1947 to 1949, resulted in the forced displacement of about 750,000 Palestinians [15,16]. Abu Lughod [4] noted that education for Palestinians during those years were immensely challenging given their dispersal across neighboring states and the lack of specifically Palestinian institutions. Palestinian themselves initiated much of the earliest educational infrastructure, even amid the chaos of ethnic cleansing. Newly displaced refugees hastily arranged classes in camps, taught by former Palestinian teachers [3]. These early schools incorporated Palestinian songs, dances, history, and geography as a means of cultural survival and identity expression [1]. Such refugee-driven activism produced the foundations that were later expanded by UNRWA after its launch. In 1949, UNRWA was created as a human development agency to provide direct relief and works programs for newly displaced Palestinian refugees in Jordan, Lebanon, Syria, and, what is now, the occupied Palestinian territories (East Jerusalem, the West Bank, and Gaza Strip) [2]. Although Western donors to UNRWA prioritized assimilating Palestinians into their host countries, Shabaneh [1] and Irfan [2,3] explained how Palestinians continued to prioritize solidifying Palestinian identity and preserving cultural knowledge across generations. Initially, UNRWA recruited Palestinian teachers as staff and used host-state curricula as expedient means for providing Palestinian education. However, the refugees successfully campaigned to reframe the educational services they were receiving to incorporate their experiences and cultural heritage. By the late 1960s, sustained activism among Palestinians led to the addition of more Palestinian historical content into textbooks. Moreover, steady enrollment progress over decades, especially at the primary level, meant wider swaths of Palestinians attained at least basic education [4]. Given this history, education holds immense significance for Palestinians in exile—both as a practical means to socioeconomic mobility in exile and, more profoundly, as a space to sustain Palestinian identity and reaffirm their right to return.

For Palestinian refugees, education has remained valued individually as a route to socioeconomic mobility but also collectively as a means of sustaining their own cultural identity as well as their social and political consciousness pertaining to their ongoing exile. However, political pressure from foreign aid donors often resulted in constraints, imbalances, and uncertainty in education for Palestinians. Donor preference for depoliticized humanitarian framings sidesteps addressing the deeper structures that subjugate Palestinian lives—such the Israeli military occupation of the occupied Palestinian territories [17,18]—thereby limiting education’s transformative potential [5]. Irfan [3] reported quality declines in the schooling of Palestinian children resulting from budget crises and funding cuts, exacerbated by the Trump administration’s 2018 decision to cut UNRWA funding for political consideration. That moment was outdone by the decision of several donor countries to cut UNRWA funding in response to allegations against UNRWA employees [19]. Over time, such external pressure transformed UNRWA operations into an education system limited to issues of access and employability rather than addressing social or political issues, thereby failing to resolve the unknowable futures of Palestinian children continuing to live under occupation or in exile [6]. This underlines the perpetual precariousness of Palestinian schooling, given its reliance on host states and humanitarian systems that prioritize avoiding thorny political issues over Palestinian humanitarian need. The Palestinian experience of education in exile epitomizes the contested politics of providing education services amid protracted conflicts, especially when beneficiaries and foreign aid donors hold conflicting visions.

It is important to recognize the context in which this study was written. At the time of writing, the Israeli military declared war on the Gaza Strip in response to attacks by the Palestinian militant groups, which resulted in over 1200 Israeli deaths on 7 October, including at least 33 children, and leading to an estimated 120 hostages [20]. The Israeli military response since 7 October resulted in over 42,000 Palestinian deaths, over a third of whom being children [20], with more than 10,000 reported missing and the death toll estimated to rise to over 100,000 [21,22]. Such actions led the International Court of Justice to declare in the case brought about by South Africa against Israel:

“In the Court’s view, the facts and circumstances mentioned above are sufficient to conclude that at least some of the rights claimed by South Africa and for which it is seeking protection are plausible. This is the case with respect to the right of the Palestinians in Gaza to be protected from acts of genocide and related prohibited acts identified in Article III, and the right of South Africa to seek Israel’s compliance with the latter’s obligations under the Convention”.

[23] (p. 54)

The war also rendered approximately 17,000 Palestinian children in the Gaza Strip as unaccompanied or separated, damaged 87% of school buildings, and left approximately 625,000 students with no access to education [20]. In light of these events and the systematic destruction of Palestinian education, continued scholarship on inclusive education for Palestinian students is needed to support the educational efforts in situations of humanitarian crises.

2.2. Conceptualizing Educational Resilience for Inclusive Education

The framing of educational resilience had been dominated by Western and neo-liberal conceptualizations of resilience that unfairly responsibilized children [24,25,26]. In contrast, this study conceptualizes educational resilience as a reflection of an interaction between the student and their wider external context rather than simply an inherent trait or internal characteristic of the student [27,28,29,30,31,32]. This is especially true for Palestinian children whose educational trajectories are directly impacted by living under illegal military occupation in the occupied Palestinian territories [17,18] or in refugee circumstances [4]. Unlike the more neoliberal conceptualizations of resilience often found in the West, Palestinians’ conceptualization of resilience is a collective one that is deeply tied to place, culture, and community. To Palestinians, Sumud (Arabic for “steadfastness” or “perseverance”) refers the preservation of cultural heritage, communal cohesion, and attachment to one’s homeland as forms of survival in the context of military occupation, forced displacement, and chronic adversity [33,34,35]. Understanding local conceptualizations of resilience is vital in inclusive education, which seeks to provide equitable learning opportunities for students from marginalized or vulnerable backgrounds like Palestinian refugees. Kiwan [36] states that inclusive education for Palestinian refugees must recognize intersecting barriers such as legal status, social class, and access to resources, underscoring the need to consider students’ complex circumstances when providing educational services. Through educational resilience as a conceptual framework, it becomes possible to identify strategies that foster not only positive educational outcomes but also the socio-emotional well-being of students, ensuring they remain engaged and motivated to learn [37]. Furthermore, research on inclusive education has shown that creating supportive school environments, where trust, belonging, and community partnerships are prioritized, plays a significant role in making education inclusive for refugee students [38,39,40]. Overall, studying educational resilience contributes to knowledge about how Palestinian students overcome the daily challenges they face and about what inclusive practices are responsive to the unique economic, social, and cultural realities of Palestinians.

Understanding the educational resilience of Palestinian students in UNRWA schools in Jordan necessitates a quantitative approach that examines their educational performance and identifies socio-ecological factors that function as resilience factors and the socio-interactional effects of these resilience factors. Adapted from Bronfenbrenner’s socio-ecological theory of learning and development, socio-ecological factors according to this study refer to the various layers of environmental influences—ranging from immediate settings like family and school to broader societal contexts—that shape a child’s educational trajectory [10,11], whereas resilience factors are socio-ecological factors that are associated with achieving positive educational outcomes and ameliorating risk [8,41]. Rutter exemplifies this perspective on educational resilience that looks at individual outcomes while recognizing environmental factors when he defines resilience as:

“Reduced vulnerability to environmental risk experiences, the overcoming of a stress or adversity, or a relatively good outcome despite risk experiences … it is an interactive concept in which the presence of resilience has to be inferred from individual variations in outcome among individuals who have experienced significant major stress or adversity”.

[14] (p. 336)

This study aims to explain the socio-interactional effects [12,13] of these resilience factors in order to consider how they contribute to the educational resilience of Palestinian students when interacting with risk. Adapted from Vygotsky’s socio-interactional theory of learning and development [10,11], socio-interactional effects according to the study refers to the type of interaction or influence specific resilience factors have on a child’s educational trajectory when interacting with a risk. The study examines two types of socio-interactional effects: promotive effects and protective effects [8,42,43]. A resilience factor has a promotive effect when it is observed to have a positive effect without the presence of risk, while a resilience factor has a protective effect when it is observed to have a positive effect in the presence of risk. Knowledge about promotive and protective effects is necessary for gaining a more contextual understanding of educational resilience processes, beyond relying on anecdotal observations of improvements that place the onus on the student. In conclusion, this study defines educational resilience as being the presence of positive educational outcomes despite exposure to significant risk or severe adversity [14,30,44], as being the result of a dynamic engagement between the student and their immediate environment [14,27,28,29], and as being specific to the socio-ecological and socio-interactional context of the student [10,11,12,13]. Understanding educational resilience is crucial for making education inclusive for Palestinian children because it recognizes how Palestinian students overcome the daily challenges they face.

2.3. Palestinians in Large-Scale Assessment Studies

Many large-scale assessment studies have examined the scholastic performance of refugee students. Some studies found that despite the challenges refugee students face, some are able to succeed academically when supported by positive school environments, teacher involvement, and targeted policies [45,46]. Other studies argued that refugee students face a significant performance disadvantage due to socio-economic and cultural barriers, and they stressed the importance of inclusive educational policies and culturally responsive teaching for refugee students’ educational trajectories [47,48]. However, many of these studies were conducted in more high-income contexts. Furthermore, focused research is still needed when examining the educational outcomes of a specific refugee group like Palestinians due to the heterogeneity of refugee populations.

Although the Israeli-occupied West Bank just recently participated in its first-ever PISA cycle, administered through the Palestinian Authority [49], this is far from being the first time that Palestinians participate. As refugees displaced in other countries, Palestinian students have been invisibilized participants in PISA and other international large-scale assessments. Since 2006, Palestinian students have participated in PISA as students of UNRWA schools in Jordan [50], but very few studies have been conducted on their scholastic performance as a population group using international large-scale assessments. Examining the scholastic performance of Palestinian students in international large-scale assessments can offer important insights into the educational trajectories of Palestinians and the factors supporting their educational achievement.

In examining PISA data, Aleghfeli [8] investigated how being an unaccompanied and separated child (UASC) impacted Palestinian refugees’ achievement in UNRWA schools in Jordan. Regression models controlling for student, teacher, and school factors found several socio-ecological factors functioning as promotive or protective for Palestinian UASCs, underscoring the importance of external factors in enabling educational resilience. They found higher economic, social, and cultural status, a positive class-disciplinary climate, and positive student–teacher relations to have promotive effects on education, meaning that Palestinian UASCs benefitted from these factors alongside other Palestinian students. They also found teachers’ use of structuring and scaffolding strategies and school ability grouping to have protective effects on education, meaning that Palestinian UASCs benefitted more from these factors than other Palestinian students.

Alongside examining data from TIMSS (Trends in International Mathematics and Science Study), another large-scale assessment, Yarrow et al. [9] conducted classroom observations and surveys comparing teaching quality in high- and low-performing West Bank schools. They found that teachers in high-performing classrooms employed more student-centered pedagogies, including discussion and debates; had higher rates of student engagement; conducted more lesson planning and outreach to parents; and made greater use of classroom resources compared to low-performing classrooms. Statistical analysis identified factors significantly correlated with classroom performance, including school culture, parental engagement, school facilities and resources, a school mission, and recognized student work. The proportion of uninvolved students was negatively correlated.

Abdul-Hamid and colleagues [7] examined both PISA and TIMSS to find Palestinian students in UNRWA schools in the West Bank, Gaza Strip, and Jordan to have consistently outperformed their counterparts in public schools. They posit that such strong performance is due to UNRWA teachers receiving better support and that the schools prioritized interactive teaching methods, creating an engaging learning environment. School autonomy, accountability, and high-quality teaching practices were also identified as crucial factors in promoting student performance. The study also highlighted the significance of collective resilience, with various actors, such as teachers, students, peers, and family members, playing important roles in supporting learning despite the adversities faced by Palestinian students.

3. Materials and Methods

Two PISA cycles were used for this study because they were the last PISA cycles to contain information on Palestinian refugees in the UNRWA school system:

• PISA 2009, which contains assessment results from 410 15-year-old Palestinian students in 13 UNRWA schools in Jordan;
• PISA 2015, which contains assessment results from 869 15-year-old Palestinian students in 28 UNRWA schools in Jordan.

PISA assesses the extent to which students have acquired, near the end of their compulsory education, key knowledge and skills that are essential for full participation in modern society. PISA does this by measuring the scholastic performance of 15-year-old students in reading, mathematics, and science. The assessments are also supplemented by detailed student-level and school-level questionnaires that are filled out by students and school principals, respectively. This makes PISA a multilevel dataset that includes data on students (Level-1) nested within schools (Level-2). The PISA sampling procedure involved the voluntary selection of participating countries. Within each country, a diverse range of schools was chosen to represent different types and regions. Students aged 15 were then randomly selected from these schools. The sampling process included stratification to ensure representation of various factors like geography and socioeconomic background. Finally, statistical weighting was applied to the collected data to adjust for any imbalances and provide results that were reflective of the entire population of 15-year-olds in each country [51,52]. Accordingly, 410 Palestinian students from 13 UNRWA schools were deemed PISA-eligible and selected to undergo the PISA assessments in 2009, and 869 Palestinian students from 28 UNRWA schools were deemed PISA-eligible and selected to undergo the PISA assessments in 2015.

3.1. Variables

For this study, the outcome variables were the reading achievement mean from PISA 2009 and the science achievement mean from PISA 2015. The student and teacher variables were selected based on the national scale reliabilities for Jordan as reported by the OECD technical reports [53,54]. Only variables with an acceptable reliability for Jordan [55,56] were included in this study.

3.1.1. Dependent Variables

Two outcome variables were examined as dependent variables for this study: reading literacy for the PISA 2009 sample and scientific literacy for the PISA 2015 sample. For the PISA 2009 cycle, reading literacy was defined by the OECD as:

“An individual’s capacity to understand, use, reflect on, and engage with written texts in order to achieve one’s goals, to develop one’s knowledge and potential, and to participate in society”.

[51] (p. 14)

Positive scores achieved in reading literacy indicate higher skills, and the OECD [53] reports reading literacy as demonstrating good internal consistency for Jordan, with a Cronbach’s $\alpha =0.91$. For the PISA 2015 cycle, scientific achievement was defined by the OECD as:

“The ability to engage with science-related issues and with the ideas of science as a reflective citizen. A scientifically literate person is willing to engage in reasoned discourse about science and technology, which requires the competencies to explain phenomena scientifically, evaluate and design scientific inquiry, and interpret data and evidence scientifically”.

[51] (p. 14)

Positive scores achieved in each outcome indicate higher skills in each domain. Positive scores achieved in scientific literacy indicate higher skills, and the OECD [54] reports scientific literacy as demonstrating good internal consistency for Jordan, with a Cronbach’s $\alpha =0.86$.

To generate student achievement estimates for the PISA assessments, the OECD utilized the plausible-value method, which uses all available information for each student to impute missing achievement estimates, thereby generating several plausible values for each student [57]. In PISA 2009, each reading outcome was represented by five plausible values (e.g., PV1READ~PV5READ) scaled on 0–600 points, while in PISA 2015, each science outcome was represented by ten plausible values (e.g., PV1SCIE~PV10SCIE) scaled on 0–600 points. It is important to note, however, that the use of plausible values means that these achievement items are not actual student test scores and cannot be used to report individual achievement. Instead, these items are only suitable for reporting population achievement or sub-population achievement [58]. Additionally, the OECD [53,54] states that survey weights must be applied to an averaging of the plausible values to generate valid parameter estimates and calculate appropriate standard errors. As such, the study uses sampling weights reported by the OECD [53] for the PISA 2009 sample and sampling weights reported by the OECD [54] for the PISA 2015 sample to ensure that the analysis calculates appropriate final estimates and standard errors. Steps to generate those final estimates and standard errors are explained in the estimation section.

3.1.2. Independent Variables

First, this study aimed to identify which learning and teaching strategies function as promotive factors. To do so, this study examined the effect of different learning and teaching strategies associated with positive reading literacy and science literacy outcomes. As such, several student and teacher variables referring to learning and teach strategies respectively were selected as independent variables for this study based on the national scale reliabilities for Jordan as reported by the OECD technical reports [53,54]. Only variables having an acceptable reliability for Jordan [55,56] were included in this study. First, both PISA 2009 and PISA 2015 contain student-reported information on learning strategies used by students. PISA 2009 includes information on six continuous variables: student enjoyment of reading, student online reading, student memorization strategies, student control strategies, and student library use. PISA 2015 includes information on five continuous variables: student enjoyment of science, student awareness of environmental matters, students’ science self-efficacy, students’ epistemological beliefs about science, and students’ instrumental motivation to learn science. Supplementary S1 (see Supplementary) provides detailed information on the student variables used in the study, including their scale reliabilities. Second, both PISA 2009 and PISA 2015 contain student-reported information on teaching strategies used by teachers. PISA 2009 includes information on four continuous variables: disciplinary climate in reading classes, teacher–student relations, teacher simulation on reading engagement, and teacher use of structuring and scaffolding strategies. PISA 2015 includes information on five continuous variables: disciplinary climate in science classes, enquiry-based teaching and learning practices, teacher support in science classes, teacher-directed science instruction, and perceived feedback from teachers. Supplementary S2 (see Supplementary) provides detailed information on the teacher variables used in the study, including their scale reliabilities. Learning and teaching strategies that are found to be positively associated with reading literacy and science literacy outcomes are deemed as promotive factors.

Second, the study aimed to identify which learning and teaching strategies function as promotive or protective factors. Kiwan [36] stresses the importance of understanding the economic, social, and cultural realities of Palestinian students when studying inclusive education for Palestinians. To do so, this study examined the extent to which the effect of different learning and teaching strategies are associated with positive reading literacy and science literacy outcomes when varied by school membership and student economic, cultural, and social status. School membership refers to an identifier variable that indicates the school to which each student belongs. Student economic, cultural, and social status refers to a scale index derived from three other variables: home possessions (wealth, cultural, and educational; books at home), higher parental occupation, and higher parental education [53,54]. Learning and teaching strategies that were found to be positively associated with reading literacy and science literacy outcomes, but not when varied by school membership and student economic, cultural, and social status were deemed as promotive factors. Contrastingly, learning and teaching strategies that were found to be positively associated with reading literacy and science literacy outcomes, including when varied by school membership and student economic, cultural, and social status were deemed as protective factors.

3.2. Analytical Strategy

3.2.1. Mixed-Effects Models

This study constructed and analyzed mixed-effects models to examine variations in the relationship between reading literacy and literacy on one the hand and the student and teacher variables on the other. Mixed-effects models are regression models that include both fixed effects and random effects, which are useful for analyzing multilevel data such as PISA [59]. Fixed effects are coefficients that are assumed to be fixed for all the students, while random effects are coefficients that vary by student. Since this study examined the impact of school membership and student economic, cultural, and social status on Palestinian students’ educational outcomes, these two variables were assigned as random effects in the model, while all other student and teacher variables were assigned as fixed effects. Cronbach [60] and Campbell [61] emphasized the importance of systematically adding sets of variables based on their theoretical relevance. This study examined two types of mixed-effects models: a random intercepts model and a random slope model [59]. In the first model, the intercept was allowed to vary randomly by school membership and student-level residuals were estimated, while the slopes remained fixed. In the second model, the intercept was allowed to vary randomly by school membership and student-level residuals were estimated, while the slope was allowed to vary randomly by student economic, cultural, and social status. The logic behind varying student economic, cultural, and social status was to better represent varying levels of risk driven by the socio-economic realities of Palestinians. A likelihood ratio test was then used to compare their goodness of fit, while an intraclass correlation coefficient (ICC) was calculated for each model to identify the proportion of the total variance in the outcome variable explained by the random effects.

3.2.2. Estimation Procedures

Given the structure of the dataset containing outcome variables made up of a set of plausible values that are multiply imputed, several estimation steps were taken that incorporated PISA’s survey weights. Procedures to pool plausible values to estimate regression coefficients and standard errors of the mixed effects models are based on Rubin’s rules [62], and survey weights were applied as recommended by the OECD [53,54]. First, the analysis was run five times, once per plausible value, to generate five parameter estimates and five sampling errors. Second, the five parameter estimates and the five sampling errors were averaged to produce the final parameter estimate and final sampling error, respectively. Third, the magnitude of the imputation error was estimated. Lastly, the final standard error was calculated by combining the sampling error and the imputation error. The mixed-effects models and the likelihood ratio tests were computed on R [63] using the mixPV and lrtPV commands, respectively, by Huang [64] and the ICC was estimated using the mixed.sdf command part of the EdSurvey package [65,66]. Supplementary S3 (see Supplementary) documents the research code built to conduct this study.

4. Findings

4.1. Resilience Factors of Reading Literacy

Table 1. Mixed-effects model of reading literacy with school membership as the random intercept.

Fixed Effects	Estimate	Standard Error	T-Statistic	df
(Intercept)	418.96 ***	10.94	38.32	93.9
Student enjoyment of reading	15.98 **	6.06	2.64	360.5
Student online reading	4.53 *	1.96	2.31	557.1
Student memorization strategies	−5.02	4.36	−1.15	63.1
Student elaboration strategies	6.28	4.36	1.44	49
Student control strategies	13.75 ***	3.76	3.65	287.6
Student library use	−8.04	5.3	−1.52	38
Teacher-student relations	9.35 *	4.67	2	76
Disciplinary climate in reading classes	4.56	4.59	1	169.4
Teacher stimulation of reading engagement	1.13	5.03	0.22	231.6
Teacher use of structuring and scaffolding strategies	−0.61	3.71	−0.17	238
Random Effects	Estimate	Standard Error	T-Statistic	df
School-level random intercept	445.69 **	158.65	2.81	1435
Student-level residuals	3880.49 ***	332.64	11.67	175

Note: * p < 0.05, ** p < 0.01, *** p < 0.001.

shows the results of the first mixed-effects model of reading literacy, which includes school membership as a random intercept. The average reading literacy score for Palestinian students in UNRWA schools in Jordan was 418.96 when all other factors were constant $(t\left(93.9\right)=38.32, p<0.001)$. However, changes to the reading literacy score were observed when changes to the student or teacher variables occurred. For every increase in student enjoyment of reading, the reading literacy score increased by 15.98 $(t\left(360.5\right)=2.64, p<0.01)$. For every increase in student online reading, the reading literacy score increased by 4.53 $(t\left(557.1\right)=2.31, p<0.05)$. For every increase in student control strategies, the reading literacy score increased by 13.75 $(t\left(287.6\right)=3.65, p<0.001)$. For every increase in teacher–student relations, the reading literacy score increased by 9.35 $(t\left(76\right)=2, p<0.05)$. In other words, student enjoyment of reading, student online reading, student use of control strategies, and teacher–student relations were associated with higher reading literacy. There was also significant variability in reading literacy scores between schools $(\beta =445.69, t\left(1435\right)=2.81, p<0.01)$ and within schools $\left(\beta =3880.49,t\left(175\right)=11.67,p<0.001\right)$. The ICC of the model was found to be 0.10, meaning that 10% of the total variance in reading literacy is attributable to differences between schools. This implies that Palestinian students’ reading literacy were significantly influenced by the specific UNRWA school they attended.

Table 2. Mixed-effects model of reading literacy with school membership as the random intercept and economic, social, and cultural status as the random slope.

Fixed Effects	Estimate	Standard Error	T-Statistic	df
(Intercept)	414.74 ***	10.86	38.19	35.3
Student enjoyment of reading	15.37 *	6.09	2.52	259.2
Student online reading	3.95	2.15	1.84	65.2
Student memorization strategies	−6.1	4.7	−1.3	98.5
Student elaboration strategies	5.38	4.21	1.28	78.7
Student control strategies	15.37 ***	4.25	3.62	642.7
Student library use	−8.56	5.04	−1.7	32.5
Teacher-student relations	8.74	4.63	1.89	85.1
Disciplinary climate in reading classes	4.25	4.79	0.89	137.4
Teacher stimulation of reading engagement	1.41	5.35	0.26	335.2
Teacher use of structuring and scaffolding strategies	−0.25	3.94	−0.06	358
Random Effects	Estimate	Standard Error	T-Statistic	df
School-level random intercept	920.2 **	287.9	3.2	106.6
Student-level random slopes	289.7 *	116.1	2.5	288.7
Student-level residuals	3628.3 ***	216.5	16.8	16.1

Note: * p < 0.05, ** p < 0.01, *** p < 0.001.

shows the results of the second mixed-effects model of reading literacy, which the adds student economic, social, and cultural status as a random slope. The average reading literacy score for Palestinian students in UNRWA schools in Jordan decreased to 414.74 when all other factors were constant $(t\left(35.3\right)=38.19, p<0.001)$. Additionally, the positive association of student enjoyment of reading with reading literacy slightly decreased to 15.37 $(t\left(259.2\right)=2.52, p<0.05)$, while the positive association of student control strategies with reading literacy increased to 15.37 $(t\left(642.7\right)=3.62, p<0.001)$. However, student online reading and teacher–student relations stopped being associated with higher reading literacy when accounting for variability in students’ economic, social, and cultural status. In short, these results reveal two promotive factors (student online reading and teacher–student relations) and two protective factors (student enjoyment of reading and student control strategies) for Palestinian students in reading literacy.

Alongside significant variability in reading literacy scores between schools $(\beta =920.2, t\left(106.6\right)=3.2, p<0.01)$ and within schools $(\beta =3628.3, t\left(16.1\right)=16.8, p<0.001)$, there was also significant variability in the effect of economic social and cultural status on reading literacy scores between schools $(\beta =289.7, t\left(288.7\right)=2.5, p<0.05)$. The ICC of the model was found to be 0.25, meaning that 25% of the total variance in reading literacy is attributable to differences between schools when accounting for student economic, social, and cultural status. This implies that some UNRWA schools were better at mitigating disadvantages associated with lower economic, social, and cultural status for Palestinian students on reading literacy. The likelihood ratio test found a statistically significant difference between the two models $(F=39.07,r\left(2,3.92\right)=4.58, p<0.05)$, justifying the addition of student economic, social, and cultural status as a random slope.

4.2. Resilience Factors of Science Literacy

Table 3. Mixed-effects model of science literacy with school membership as the random intercept.

Fixed Effects	Estimate	Standard Error	T-Statistic	df
(Intercept)	412.35 ***	6.22	66.32	780.8
Student enjoyment of science	4.01	2.68	1.5	3253.8
Student awareness of environmental matters	12.4 ***	2.4	5.18	606.9
Student science self-efficacy	3.47	2.49	1.4	158.4
Student epistemological beliefs about science	10.32 **	3.65	2.83	73.8
Student instrumental motivation for science	5.05	3.31	1.53	460
Disciplinary climate in science classes	3.69	3.32	1.11	90.7
Enquiry-based science teaching and learning practices	−7.44	5.04	−1.48	790.1
Teacher support in science classes	−0.58	4.16	−0.14	182.2
Teacher-directed science instruction	3.56	2.42	1.47	47.5
Perceived feedback from teacher	−13.24 **	4.12	−3.22	357.1
Random Effects	Estimate	Standard Error	T-Statistic	df
School-level random intercept	695.42 **	225.5	3.08	187.7
Student-level residuals	3724.95 ***	255.38	14.59	97.8

Note: ** p < 0.01, *** p < 0.001.

shows the results of the mixed-effects model of science literacy when adding school membership as a random intercept. The average science literacy score for Palestinian students in UNRWA schools in Jordan was 412.35 when all other factors were constant $(t\left(780.8\right)=66.32, p<0.001)$. However, changes to the science literacy score were observed when changes to the student or teacher variables occurred. For every increase in student awareness of environmental matters, the science literacy score increased by 12.4 $(t\left(606.9\right)=5.18, p<0.001)$. For every increase in student epistemological beliefs about science, the science literacy score increased by 10.32 $(t\left(73.8\right)=2.83, p<0.01)$. However, for every increase in perceived feedback from teacher, the science literacy score decreased by 13.24 $(t\left(357.1\right)=-3.22, p<0.01)$. In other words, student awareness of environmental matters and student epistemological beliefs about science were associated with higher science literacy. There was also significant variability in science literacy scores between schools $(\beta =695.43, t\left(187.7\right)=3.08, p<0.01)$ and within schools $(\beta =3724.95, t\left(97.8\right)=14.59, p<0.001)$. The ICC of the model was also found to be 0.16, meaning that 16% of the total variance in science literacy is attributable to differences between schools. This implies that Palestinian students’ science literacy was significantly influenced by the specific UNRWA school they attended.

Table 4. Mixed-effects model of science literacy with school membership as the random intercept and economic, social, and cultural status as the random slope.

Fixed Effects	Estimate	Standard Error	T-Statistic	df
(Intercept)	414.91 ***	5.87	70.66	513.6
Student enjoyment of science	4.11	2.62	1.57	4692.1
Student awareness of environmental matters	11.7 ***	2.31	5.06	310.1
Student science self-efficacy	3.45	2.47	1.4	155
Student epistemological beliefs about science	11.18 **	3.64	3.07	74.2
Student instrumental motivation for science	4.73	3.25	1.46	313.8
Disciplinary climate in science classes	2.62	3.24	0.81	79.5
Enquiry-based science teaching and learning practices	−7.81	5.01	−1.56	664.4
Teacher support in science classes	0.59	4.29	0.14	142.1
Teacher-directed science instruction	3.58	2.45	1.46	43.6
Perceived feedback from teacher	−13.9 ***	4.14	−3.36	537.2
Random Effects	Estimate	Standard Error	T-Statistic	df
School-level random intercept	731.81 **	227.38	3.22	163.6
Student-level random slopes	254.3 **	96.89	2.62	233.4
Student-level residuals	3501.55 ***	160.15	21.86	16.1

Note: ** p < 0.01, *** p < 0.001.

shows the results of the mixed-effects model of science literacy when adding student economic, social, and cultural status as a random slope. The average science literacy score for Palestinian students in UNRWA schools in Jordan increased to 414.91 when all other factors were constant $(t\left(513.6\right)=70.66, p<0.001)$. Additionally, the positive association of student awareness of environmental matters with science literacy score slightly decreased to 11.7 $(t\left(310.1\right)=5.06, p<0.001)$, while the positive association of student epistemological beliefs about science with science literacy slightly increased to 11.18 $(t\left(74.2\right)=3.07, p<0.01)$. Moreover, the negative association of perceived feedback from teacher with science literacy slightly increased to 13.9 $(t\left(537.2\right)=-3.36, p<0.001)$. In short, these results reveal two protective factors (student awareness of environmental matters and student epistemological beliefs about science) for Palestinian students in science literacy.

Alongside significant variability in science literacy scores between schools $(\beta =731.81, t\left(163.6\right)=3.22, p<0.01)$ and within schools $(\beta =3501.55, t\left(16.1\right)=21.86, p<0.001)$, there was also significant variability in the effect of economic social and cultural status on science literacy scores between schools $(\beta =254.3, t\left(233.4\right)=2.62, p<0.01)$. The ICC of the model was found to be 0.21, meaning that 21% of the total variance in science literacy is attributable to differences between schools when accounting for student economic, social, and cultural status. This implies that some UNRWA schools were better at mitigating disadvantages associated with lower economic, social, and cultural status for Palestinian students on science literacy. The likelihood ratio test found a statistically significant difference between the two models $(F=52.92, t\left(2,1575\right)=2.14, p<0.001)$, justifying the addition of student economic, social, and cultural status as a random slope.

5. Discussion of Findings

Table 5. Promotive and protective factors of reading literacy and science literacy for Palestinian students.

Factors	Reading Literacy	Science Literacy
Promotive	Student online reading Teacher–student relations
Protective	Student enjoyment of reading Student control strategies	Student awareness of environmental matters Student epistemological beliefs about science

showcases the different promotive and protective factors identified by the study. First, two promotive factors were identified that had positive associations with reading literacy but whose positive association decreased significantly with higher economic, social, and cultural status: student online reading and teacher–student relations. Regarding student online reading, this finding should not be surprising, as access to the internet is not a societal norm but a privilege resulting from higher socio-economic status, specifically for refugee populations. Refugees are 50% less likely than the global non-refugee population to have an Internet-enabled phone [67]. Regarding teacher–student relations, this finding could imply that teacher–student relations were stronger for students with high socio-economic status and weaker for students with low socio-economic status. Several studies have shown that teachers perceive students from higher socio-economic backgrounds better than students from lower socio-economic backgrounds [68,69].

Second, two protective factors were identified that had positive associations with reading literacy regardless of economic, social, and cultural status: student enjoyment of reading and student control strategies. Regarding student enjoyment of reading, this finding reaffirms the importance of cultivating the joy of reading among Palestinian students, especially in the early years. Reading for pleasure initiated in early childhood was found to be associated with better cognitive performance, mental well-being, and brain structure in young adolescence [70]. Regarding student control strategies, the findings could reflect the importance of self-regulated learning practices, as reflected in students’ ability to control their own metacognition, cognition, motivation, affect, and behavior [71], in enhancing reading performance regardless of socio-economic background.

Third, two protective factors were identified that had positive associations with science literacy regardless of economic, social, and cultural status: student awareness of environmental matters and student epistemological beliefs about science. Regarding student awareness of environmental matters, this finding warrants the importance of integrating environmental education into traditional science classrooms. Teaching and learning about environmental matters help students identify, apply, and analyze abstract scientific concepts to their environmental surroundings, thereby increasing their scientific knowledge [72]. Regarding student epistemological beliefs about science, this finding reaffirms the need to emphasize epistemic reasoning in science education (e.g., “What is a theory?,” “What is a hypothesis?,” “What is scientific evidence?”) and not just content or procedural knowledge about scientific concepts and constructs [73,74].

This study reports some methodological limitations that should be carefully considered. First, the use of a cross-sectional design for secondary data analysis prevented the establishment of causality. This is often the challenge with secondary data analyses because they involve the use of existing quantitative data to find answers to a research question that is different from the original work [75]. For example, the PISA dataset was not designed with Palestinian refugees in mind due to it originating from Western models of education [76]. Accordingly, future research should consider longitudinal studies that capture quantifiable educational data on Palestinian children to allow for a more robust quantitative examination of which resilience factors (such as those identified in this study) are salient in predicting change in educational outcomes. Another limitation of the study is that it relied on student-reported responses for the variables, as the use of self-reported responses may introduce social desirability bias or recall bias. Regardless, the study delivered its best efforts to produce accurate parameter estimates by committing to a strict estimation approach as outlined by the OECD [53,54]. While the study provides valuable insights, its limitations, including the cross-sectional design and reliance on the PISA dataset, should be considered when interpreting the findings.

6. Conclusions and Implications

The study makes several important contributions to knowledge about the educational needs of Palestinian children. Knowledge of promotive and protective factors provide valuable insights into correlates that may contribute to positive educational outcomes for Palestinian students in the face of risk and adversity. There are opportunities for future quantitative experimental and quasi-experimental designs to investigate the effect of one specific promotive or protective factor on Palestinian students’ educational outcome. Given that this study was done only on Palestinian students in Jordan, future researchers can expand this study to investigate teaching and learning strategies that contribute to Palestinian students’ educational resilience in the occupied Palestinian territories as well as in the diaspora. Moreover, this study reaffirms the need to study Palestinian children’s abilities to navigate and negotiate with their surrounding ecosystem to overcome challenges to their educational trajectories, which is key to understanding their educational resilience [77]. In the context of research on inclusive education and refugee education, this paper calls for more qualitative and case-selective designs that explore the learning experiences of Palestinian children in contexts of military occupation and forced displacement. Lastly, this study reminds researchers of the importance of centering Palestinians’ experience when researching their educational trajectories by considering the unique adversities faced by Palestinian children, rather than through the usual decontextualizing of Palestinian education. Palestinians’ culture, national identity, and place in the world must be recognized and preserved considering their historical experience and recent events at the time of writing, and the academic community can play such a role by incorporating Palestinian narratives in studies about and for Palestinians.

The findings of this study carry important implications for reading and science educators working with Palestinian children, whether in the occupied Palestinian territories or in countries that have or expect to receive forcibly displaced Palestinian refugees. In the context of reading education, creating a classroom environment that fosters students’ love of reading—such as by incorporating Palestinian literature into curriculums—and encourages student use of control strategies, such as goal setting and planning study sessions, may enhance Palestinian students’ reading literacy, even in the face of adversity. This aligns with research on inclusive education for refugees, which emphasizes the need for holistic approaches that address not only academic content but also the socio-emotional well-being of refugee students [8,37]. In the context of science education, the study’s findings highlight the importance of fostering students’ awareness of environmental issues and encouraging positive epistemological beliefs about science. This can be achieved by learning about the flora and fauna of their homeland, ecological concerns associated with war and conflict, and sustainable practices that promote environmental conservation. Palestinian students can also greatly benefit from learning that prioritizes scientific inquiry and evidence-based reasoning, rather than rote learning of scientific concepts [73,74]. As the literature on refugee education has shown, pedagogical strategies that integrate refugee cultural knowledge are crucial for fostering refugee student engagement [38,39,40]. Inclusive teaching practices that recognize the educational resilience—or Sumud—of Palestinian students can empower them to thrive not just academically and individually but also for the benefit of their families and communities.