Session Room: RUU D104 HELENA
Session Moderator: Kadriye Ercikan, Senior Vice President, ETS Global Research 

This session brings together four papers that examine how adult skills function in Nordic and Baltic countries. Each paper takes a different angle on a shared question: what do we learn when we look beyond credentials and test scores to understand how adults use, perceive, and build on their competencies? The first paper examines whether adults accurately assess their own skills by comparing perceived and actual performance. The second traces how foundational literacy supports problem solving as people age. The third investigates whether skilled adults without formal credentials still find their way into further learning. The fourth examines differences in engagement in formal, non-formal, and informal learning within and between skilled and low-skilled adults in Finland. Together, they use PIAAC data to provide a more comprehensive view of adult skills systems—linking proficiency, self-perception, and learning engagement—and to argue that policy should focus not only on formal qualifications, but on what adults can do, how they understand their abilities, and how they continue to develop them over time.

---

The Nexus Between Skills and Confidence in Skills in the Nordic-Baltic Countries
Usama Ali, Principal Measurement Scientist, ETS Research Institute
Irwin Kirsch, Ralph W. Tyler Chair in Large-Scale Assessment, ETS, Emeritus 

This study examines the relationship between adults’ proficiency in literacy and numeracy and their confidence in their performance using data from the Programme for the International Assessment of Adult Competencies (PIAAC). For the first time, PIAAC Cycle 2 collected self-reported measures of perceived performance—how well respondents believed they performed on the assessment—as well as the level of effort they reported exerted, and notably elicited these judgments after completion of the cognitive assessment. This post-assessment design provides a more direct basis for comparing perceived and actual performance. We leverage these measures to assess the alignment between perceived and actual performance, operationalized through proficiency levels. 

Building on a substantial body of research on metacognitive calibration and self-assessment accuracy, which shows that individuals often misjudge their own performance, this study examines the extent to which such patterns emerge in adult skills data. Using a two-way classification that links perceived performance to actual proficiency, we identify patterns of under- and over-confidence. Individuals who report low perceived accuracy despite demonstrating higher proficiency are classified as under-confident, while those reporting high perceived accuracy with lower proficiency are considered over-confident. 

The analysis explores these patterns across Nordic-Baltic countries and examines variation by domain and demographic characteristics. Given that a substantial share of adults—often around 30 percent or more—perform below proficiency Level 3, patterns of over-confidence among lower-performing individuals may have important implications. The findings contribute to the literature by extending evidence on confidence calibration to adult populations in large-scale assessments. From a policy perspective, systematic miscalibration may affect human capital development: under-confidence may reduce engagement in lifelong learning and limit skill utilization, while over-confidence may lead to suboptimal educational investments and labor market decisions. These results highlight the importance of incorporating non-cognitive constructs, such as confidence and self-assessment accuracy, into the interpretation of PIAAC results and the design of policies aimed at improving adult skills and workforce adaptability.

---

How Literacy Shapes Problem Solving as Populations Age: Evidence from Nordic and Baltic Countries 
Andrew McEachin, Senior Research Director, Policy Research Center, ETS 

This study examines how literacy relates to problem-solving proficiency among adults in Nordic and Baltic countries, using Cycle 2 of the Programme for the International Assessment of Adult Competencies (PIAAC). We ask three questions: how problem-solving proficiency varies with literacy across the full distribution of adult skills; how this relationship differs within and between countries in the region; and how it looks different for younger versus older adults. 

We take a distributional approach rather than comparing means or proficiency-level shares. We model problem-solving performance at each point in the literacy distribution, and we let the shape of that relationship vary by country and by age group. This approach shows where along the literacy distribution the problem-solving gradient is steepest, and which adults depend most on literacy to navigate information-rich tasks. Nordic and Baltic countries share strong basic-schooling traditions but differ in labor market structure, migration, and adult learning systems. Comparing them separates patterns specific to one country from patterns shared across the region. 

PIAAC Cycle 2 covers working-age adults across several decades of birth cohorts. This lets us compare younger and older respondents within the same survey. We ask whether adults with weaker literacy show a wider problem-solving gap between older and younger respondents, and whether adults with stronger literacy show a narrower one. This is a cross-sectional comparison, not a within-person trajectory. Even so, it shows how literacy and adaptive competence line up across the adult population a country has today. 

PIAAC already identifies who has low literacy. The contribution here is to show how literacy gaps constrain the adaptive, information-rich problem solving that labor markets increasingly demand, and how that constraint looks different across age groups. The findings bear directly on lifelong learning. If problem solving rests heavily on literacy, adult learning systems, workforce training, and upskilling initiatives cannot assume a floor of basic skills that many adults have not reached. The results also point forward: absent change in primary and secondary schooling, today's adult literacy distribution previews tomorrow's adult problem-solving capacity. By showing where foundational literacy most constrains adaptive competence, the paper identifies concrete leverage points for policy that would widen who keeps learning and adapting as Nordic and Baltic populations age.

---

Skills Without Credentials: Adult Learning Participation Across the Life Course in Nordic and Baltic Countries  
Anita Sands, Lead Policy Research Analyst, Policy Research Center, ETS 

Adult participation in lifelong learning remains a central policy priority across OECD countries, yet recent evidence points to stagnation in participation rates and persistent inequalities across population groups. These disparities are particularly pronounced among older adults and other underrepresented groups. While credential level is consistently one of the strongest predictors of participation in adult learning, and prior research highlights the role of assessed skills, far less is known about whether participation reflects individuals' assessed skill capacity when it diverges from formal qualifications. 

Using PIAAC Cycle 2 data from Nordic and Baltic countries, we examine whether participation in adult learning reflects demonstrated proficiency or continues to track formal credentials. We focus in particular on adults who perform at Level 3 or above in literacy or numeracy despite holding lower formal qualifications and compare their participation patterns to those of higher-credentialed adults with weaker assessed skills. Drawing on descriptive and multivariate approaches, the analysis considers participation in formal and non-formal learning, job-related training, employer sponsorship, and unmet demand. The Nordic and Baltic countries vary in adult learning infrastructure and institutional context, providing comparative leverage for examining whether credential-based sorting operates similarly across different systems. 

The paper asks whether adult learning participation is more closely aligned with individuals' foundational skills or with formal credentials, and whether credential-based sorting creates misalignment between adults' capacity and their access to learning opportunities. We then examine age-group differences in these dynamics given documented declines in employer investment and participation with age. 

If participation tracks credentials more than demonstrated skills, current approaches may overlook a substantial group of adults and reinforce existing inequalities while limiting the efficiency of investment in adult learning - with implications that extend beyond this regional context.

---

Patterns of learning engagement among high- and low-performing adults in Finland
Joonas Mannonen, National Project Manager (PIAAC Finland), Finnish Institute for Educational Research, University of Jyväskylä, Finland
Kari Nissinen, Vice-Director, Finnish Institute for Educational Research, University of Jyväskylä, Finland 
Raija Hämäläinen, Professor, Department of Education, Faculty of Education and Psychology, University of Jyväskylä, Finland 
Bram De Wever, Professor, Department of Educational Studies, Faculty of Psychology and Educational Sciences, Ghent University, Belgium 

Evidence from the second cycle of the Survey of Adult Skills (PIAAC CY2) indicates that, on average, engagement in different forms of learning is a strong predictor of proficiency in information‑processing skills. At the same time, learning trajectories do not correspond linearly with proficiency levels: not all skilled adults are highly educated or actively engaged in learning, and not all low-skilled adults are disengaged from learning altogether. Moreover, the learning trajectories that most effectively support skill development may differ across population segments according to background characteristics. This complexity suggests that effective skills policy cannot rely on uniform assumptions about learning and proficiency. Instead, it requires a more nuanced understanding of participation patterns in formal, non-formal, and informal learning across proficiency levels. 

This paper responds by mapping common patterns of engagement in formal, non-formal, and informal learning among skilled adults in Finland. In parallel, it analyses patterns of limited participation and gaps in learning that are commonly found among low-skilled adults. By examining mismatches between learning engagement and skill outcomes, the study seeks to identify policy-relevant leverage points for reducing skill inequalities that emerge early in life and accumulate throughout adulthood. 

We employ Finnish PIAAC CY2 data and hierarchical cluster analysis. Skilled adults were defined as those scoring at the two highest proficiency levels in each of the three skill domains assessed in PIAAC CY2—literacy, numeracy, and adaptive problem solving (n = 1,121). Conversely, low-skilled adults were defined as those scoring at the three lowest levels in each domain (n = 825). To identify dominant patterns of learning engagement among skilled and low-skilled adults, we use educational attainment, participation in non‑formal adult education and training, parental education, learning at work and skill use at and outside work as clustering variables. A broad range of background characteristics is then employed to describe the resulting clusters and to analyze differences within and between the groups of skilled and low‑skilled adults. 

The findings contribute new insights into the lifelong development of information‑processing skills in Finland and offer implications for policymakers designing national lifelong learning strategies that effectively reach diverse segments of the adult population.