1 Introduction
1.1 Conceptual framework: criteria of academic resilience and underlying norms
1.1.1 Resilience and academic resilience
1.1.2 Measuring adversity: composite vs. distinct measures of student background
1.1.3 Measuring positive adaptation: selecting an Indicator of student outcome
1.1.4 Thresholds for adversity and positive adaptation: cross-country vs. within-country
1.2 State of research
1.2.1 Overview about academic resilience studies in international large-scale assessments
Type of threshold | Reference | Data | Student background indicator (threshold for defining adversity) | Outcome (threshold for defining positive adaptation) | Domain, Country samples | Main findings | |
---|---|---|---|---|---|---|---|
Fixed background, fixed outcome | Direct | Erberer et al. (2015) | TIMSS 2011 | HER index (“few resources”) | Fix (≥ 475) | Mathematics (Grade 8), 28 education systems | Environments of high academic achievement appear to support academic resilience. Students’ high educational aspirations appear to be the strongest and most consistent predictor of academic resilience. |
Sandoval-Hernández and Bialowolski (2016) | TIMSS 2011 | HER index (“few resources”) | Fix (≥ 475) | Mathematics (Grade 8), 5 Asian education systems | No consistent patterns were identified, student expectations and amount of time spent on homework were significantly associated with academic resilience in Singapore, boys tended to be more resilient in Korea than girls. | ||
Frempong et al. (2016) | TIMSS 2012 | New SES index based on 18 assets listed in student questionnaire: disadvantaged students (bottom 2 SES quantiles) from disadvantaged schools (bottom 3 quantiles) | Fix (≥ 352 as the South African country mean) | Mathematics (Grade 9), South Africa | Girls who speak their native language tended to be more resilient. Positive attitude, valuing learning, job aspiration were important characteristics associated with academic resilience. | ||
Fixed background, relative outcome | Sandoval-Hernandez and Cortés (2012) | PIRLS 2006 | new SES index based on parents’ education, occupation, home possessions (bottom 20% across countries in a cluster) | Within country (top 20%) | Reading, cluster of countries with comparable SES index | Non-cognitive characteristics (self-confidence and motivation) were the most important factors in predicting academic resilience. Material support provided by family and safety in school also mattered. | |
Relative background, fixed outcome | OECD (2011) | PISA 2009 | ESCS (bottom 1/3 within country) | Across countries (top 1/3) | Science all participating countries | Academic resilience was not identified as a domain-specific characteristic. No gender gap in academic resilience was found for science. Language and immigrant background were marginally associated with academic resilience in a few countries. | |
OECD (2017) | PISA 2015 PISA 2006 | ESCS (bottom 1/4 within country) | Across countries (top 1/4) | Science, all participating countries | In countries with a large increase in the proportion of academically resilient students, the percentage of low-performing students was reduced while the average performance was maintained or improved. | ||
OECD (2018) | PISA 2015 | ESCS (bottom 1/4 within country) | Across countries (top 1/4) | Science, all participating countries | The presence of academic resilience varies greatly in this operationalization, 76% of disadvantaged students are academically resilient in Viet Nam, but less than 5% in Kosovo, Peru, and Tunisia. | ||
García-Crespo et al. (2019) | PIRLS 2016 | New SES index based on possessions and books at home, highest academic qualifications, and level of employment of parents (bottom 25% within countries) | Across EU countries (top 25%) | Reading, EU countries that participated in this study | Student confidence in reading and a favorable school climate greatly increased the likelihood of academic resilience. | ||
Residual | OECD (2010) | PISA 2009 | ESCS (bottom 1/4 within country) | Across countries (top 1/4 of residual) | Reading, all participating countries | Academically resilient students were more prevalent in those education systems that PISA indicators showed to be more equitable. | |
Cheung et al. (2014) | PISA 2009 | ESCS (bottom 1/4 within country) | Across countries (top 1/4 of residual) | Reading, 4 East Asian economies | Family structure, expected education, kindergarten attendance, and reading engagement were associated with academic resilience. | ||
Agasisti and Longobardi (2014) | PISA 2009 | School ESCS and student SES (bottom 1/3 students from bottom 1/3 schools within country) | Across countries (top 1/3 of residual) | Reading, Italy | Motivation and extracurricular activities were positively associated with academic resilience. Immigrant status was negatively associated with academic resilience. | ||
Agasisti and Longobardi (2014) | PISA 2000–2012 | School ESCS and student SES (bottom 1/3 students from bottom 1/3 schools within country) | Across countries (top 1/3 of residual) | Reading, 58 education systems over period 2000–2012 | Educational funding can help disadvantaged students in overcoming their penalizing starting conditions. | ||
Cheung (2017) | PISA 2012 | ESCS (bottom 1/4 within country) | Across countries (top 1/4 of residual) | Mathematics, Five East Asian education systems | Self-efficacy was strongly, familiarity with mathematical concepts learned in the earlier grades moderately, and anxiety weakly associated with academic resilience. | ||
Agasisti and Longobardi (2017) | PISA 2009 | School ESCS and student SES (bottom 1/3 students from bottom 1/3 schools within country) | Across countries (top 1/3 of residual) | Reading, 5 EU countries | More extracurricular activities and a more positive school climate were positively associated with academic resilience. | ||
Across-domain | Agasisti et al. (2018) | PISA 2006–2015 | ESCS (bottom 1/4 within country) | ≥ “Level 3” in all three domains | Reading, mathematics and science, all participating countries | Classroom disciplinary climate was positively associated with academic resilience. The amount of human and material resources available in school was weakly associated with academic resilience. | |
OECD (2018) | PISA 2015 | ESCS (bottom 1/4 within country) | ≥ “Level 3” in all three domains | Reading, mathematics and science, all participating countries | This type of cross-domain academic resilience was more frequently observed in countries with higher average performance. Disciplinary climate and motivation were associated with academic resilience. | ||
Non-cognitive | OECD (2018) | PISA 2015 | ESCS (bottom 1/4 within country) | Socially and emotionally satisfied | Non-cognitive, 48 countries | Students who were socially and emotionally resilient also tended to do better academically. Lower shares of socio-emotionally resilient students were found in the top-performing systems than with the application of a cognitive outcome definition of resilience. | |
Relative background, relative outcome | OECD (2011) | PISA 2009 | ESCS (bottom 1/3 within country) | Within country (top1/3) | Science, all participating countries | The extent to which students adopted positive approaches to learning and the amount of time they spent in regular science lessons were strongly associated with academic resilience. | |
Karklina (2012) | PISA 2006 interview data | ESCS (bottom 1/3 within country) | Within country (top1/3) | Science, Latvia | Parental education level and length of poverty experience were associated with academic resilience. | ||
Aydiner and Kalender (2015) | PISA 2012 | ESCS, bottom 1/4 within country | Within country (top1/3) | Reading, Turkey | Sense of belonging was associated with academic resilience | ||
OECD (2018) | PISA 2015 | ESCS, bottom 1/4 within country | top1/4 within countries | Science, all participating countries | Academic resilience was not more frequently observed in top-performing countries than in developing or low-performing countries. |
1.2.2 Fixed background and fixed outcome thresholds
1.2.3 Fixed background and relative outcome thresholds
1.2.4 Relative background and fixed outcome thresholds
1.2.5 Relative background and relative outcome thresholds
1.3 Does the conceptualization of academic resilience matter: a question of validity
2 The present study
3 Methods
3.1 Sample
3.2 Measures
3.2.1 Disadvantaged student background: Adversity
3.2.2 Strong educational outcomes: positive adaptation
3.2.3 Validity measures
3.3 Data analysis
4 Results
4.1 How do different conceptualizations of academic resilience affect how much and which students are classified as academically resilient?
4.1.1 First step: defining disadvantaged students (adversity)
Norway | Peru | Hong Kong | |
---|---|---|---|
ESCS | 7.18% | 63.94% | 46.76% |
WEALTH | 2.51% | 78.51% | 55.66% |
BOOKS | 20.80% | 65.77% | 36.41% |
EMOSUPS | 32.86% | 40.14% | 63.30% |
4.1.2 Second step: defining well-performing students (positive adaptation)
4.1.3 Third step: conceptualizing academically resilient students by combining background and performance definitions
Proportion of disadvantaged students classified as resilient | Proportion of all students classified as resilient | |||||
---|---|---|---|---|---|---|
Norway | Peru | Hong Kong | Norway | Peru | Hong Kong | |
Fixed ESCS | 41.58% | 9.85% | 73.18% | 2.99% | 6.30% | 34.22% |
Relative ESCS | 50.06% | 4.71% | 70.71% | 16.16% | 1.56% | 23.07% |
Proportion of disadvantaged students classified as resilient | Proportion of all students classified as resilient | |||||
---|---|---|---|---|---|---|
Norway | Peru | Hong Kong | Norway | Peru | Hong Kong | |
Fixed ESCS | 13.52% | 22.75% | 28.93% | 0.97% | 14.55% | 13.53% |
Relative ESCS | 20.89% | 13.17% | 26.66% | 6.75% | 4.37% | 8.70% |
4.2 Which students are classified as academically resilient in the different conceptualizations?
Back | Out | Norway | Peru | Hong Kong | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Female | Male | Female | Male | Female | Male | ||||||||
n | % | n | % | n | % | n | % | n | % | n | % | ||
Fix | Fix | 94 | 1.72 | 69 | 1.26 | 159 | 2.28* | 280 | 4.02 | 905 | 16.89 | 929 | 17.34 |
Rel | 33 | 0.60 | 20 | 0.37 | 427 | 6.13* | 587 | 8.42 | 324 | 6.05* | 401 | 7.48 | |
Rel | Fix | 463 | 8.49 | 419 | 7.68 | 31 | 0.44* | 78 | 1.12 | 608 | 11.35 | 628 | 11.72 |
Rel | 190 | 3.48 | 178 | 3.26 | 120 | 1.72* | 185 | 2.65 | 210 | 3.92* | 256 | 4.78 |
Back | Out | Norway | Peru | Hong Kong | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Diff L | Same L | Diff L | Same L | Diff L | Same L | ||||||||
n | % | n | % | n | % | n | % | n | % | n | % | ||
Fix | Fix | 31 | 0.57* | 132 | 2.42 | 7 | 0.10* | 432 | 6.20 | 26 | 0.49* | 1808 | 33.74 |
Rel | 5 | 0.09* | 48 | 0.88 | 21 | 0.30* | 993 | 14.24 | 2 | 0.04* | 723 | 13.49 | |
Rel | Fix | 97 | 1.78* | 785 | 14.39 | 4 | 0.06* | 105 | 1.51 | 21 | 0.39* | 1215 | 22.67 |
Rel | 33 | 0.60* | 335 | 6.14 | 13 | 0.19* | 292 | 4.19 | 2 | 0.04* | 464 | 8.66 |
4.3 How are different conceptualizations associated with external variables?
Norway ESCS | Peru ESCS | Hong Kong ESCS | |||||
---|---|---|---|---|---|---|---|
Fixed | Relative | Fixed | Relative | Fixed | Relative | ||
Performance | Fixed | 1.05 | 1.00 | 1.30* | 1.39* | 0.95 | 0.97 |
Relative | 1.02 | 0.99 | 1.35* | 1.40* | 0.94 | 0.92 |
Norway | Peru | Hong Kong | |||||
---|---|---|---|---|---|---|---|
ESCS | ESCS | ESCS | |||||
Fixed | Relative | Fixed | Relative | Relative | Relative | ||
Performance | Fixed | 0.72* | 0.70* | 0.94 | 1.00 | 0.69* | 0.72* |
Relative | 0.75* | 0.71* | 0.98 | 1.04 | 0.68* | 0.69* |