This chapter describes key aspects of measuring informal learning, including how it is captured in international surveys such as the Adult Education Survey (AES) and OECD Survey of Adult Skills (PIAAC). It also examines national approaches, explores differences in participation rates across surveys, expands on the challenge with self-reported information on informal learning, and considers recent efforts to move beyond the traditional triadic classification of learning types. It concludes by presenting six OECD recommendations aimed at strengthening international survey instruments (Priority Topic 3), complementing international instruments with national measures (Priority Topic 4), and promoting ongoing innovation in the definition and measurement of informal learning (Priority Topic 5).
Giving Informal Learning the Recognition it Deserves
Abstract
Assessment
The AES is the main cross-country survey in the European Union (EU) for measuring informal learning at the individual level. Variations in questionnaire design, translation, and national implementation have led to cross-country inconsistencies, although consistency has increased since 2016. Recent updates also reflect evolving practices, such as the 2022 inclusion of electronic devices.
The OECD Survey of Adult Skills (PIAAC) captures workplace‑related informal learning (e.g. learning new things, learning-by-doing, and staying updated). It provides detailed frequency measures, but it is limited to job-related contexts.
Participation rates vary considerably across surveys, reflecting differences in definitions, response formats, data collection methods, and survey populations.
Many surveys rely on respondents’ ability to recognise and interpret examples, risking underestimation and cross-country variation, especially among groups less familiar with the terminology.
In most OECD countries, informal learning is measured mainly through international frameworks such as AES and PIAAC, which enable comparability across countries.
A smaller group of countries, such as the Netherlands and Australia, complement these with national surveys tailored to their policy priorities. These provide more granular data on workplace and lifelong learning practices.
Priority Topic 3: Strengthening international survey instruments
Several international surveys measure informal learning across different contexts, but the overall framework remains fragmented and inconsistent, with limitations in conceptual alignment, indicator coverage, comparability over time, timeliness of data, and cross-country implementation.
Recommendation 5: Preserve and extend the measures of informal learning included in international survey instruments. Retain and extend international survey measures to better capture participation, intensity, outcomes, and emerging forms.
Recommendation 6: Promote greater consistency in the implementation of international surveys at the national level. Support comparability by applying harmonised definitions, guidance, and implementation practices across countries.
Priority Topic 4: Complementing international instruments with national measures
Many countries already measure informal learning in national surveys, complementing international surveys by offering greater frequency, depth, and flexibility to address country-specific priorities. However, varying definitions, survey design, and question formats, limit comparability across these instruments.
Recommendation 7: Include measures of informal learning in national survey instruments. Use national surveys to complement international data and address country-specific evidence and policy needs.
Recommendation 8: Establish an international Informal Learning Expert Working Group. Create a dedicated platform for international collaboration on definitions, measurement, and methodological development.
Priority Topic 5: Promoting ongoing innovation in the definition and measurement of informal learning
Informal learning can be measured not only through surveys but also via innovative methods such as time‑use diaries, ethnographic studies, social network analysis, and learning analytics. As informal learning changes with technology and changing work patterns, its definitions and measurement approaches must be continually revisited to reflect its dynamic nature.
Recommendation 9: Identify informal learning measurement as an explicit research priority. Support research and testing of innovative approaches to improve the measurement and understanding of informal learning.
Recommendation 10: Regularly review and update national and international measurement instruments. Ensure definitions and measurement tools remain relevant as learning practices and technologies evolve.
Understanding how informal learning is measured is essential for accurately assessing adult learning and education policies. Measurement approaches can vary significantly across countries and surveys, and these differences can significantly affect reported participation rates and their interpretation. As a result, they have important implication for international comparisons and the formulation of evidence‑based policy.
Measurement of informal learning in international surveys
Informal learning is measured in several international surveys that enable cross-country comparisons. The two most prominent are the AES and PIAAC, with the latter focussing solely on informal learning in the workplace. These two surveys are the primary focus of this section.
Other surveys also capture elements related to informal learning, though often in ways that do not fully align with standard definitions. For example, the EU Continuing Vocational Training Survey (CVTS) includes indicators on learning through job rotation, conferences and workshops, and self-directed learning. However, the CVTS requires these activities to be planned, which diverges from widely accepted definitions of informal learning that emphasise its unstructured and non-institutional nature (see also the assessment in Chapter 3). The EU Labour Force Survey (LFS) does not include questions specifically on informal learning, but only on participation in formal and non-formal education or training, as defined by the CLA. UNESCO and their Global Report on Adult Learning and Education (GRALE) also acknowledges informal learning, but monitors informal learning indirectly through country-reported policy measures, particularly recognition of prior learning (RPL), workplace and community-based learning opportunities, and the integration of informal learning into lifelong learning systems.
The European Centre for the Development of Vocational Training (CEDEFOP), in collaboration with Eurofound, has measured some of the conditions for informal learning in the workplace through its European Company Survey (ECS). The latest ECS was conducted in 2019 and explored whether employees have opportunities to learn new things and whether managers view learning from colleagues or supervisors, or learning by doing as important (Eurofound / Cedefop, 2019[1]).
More recently, CEDEFOP conducted the European Training and Learning Survey (ETLS), an EU-wide survey seeking to better understand learning for work, including questions exploring participation in, motivation for, and outcomes of informal learning activities (CEDEFOP, 2026[2]) (Verian, 2024[3]) (see Box 4.1).
The European Training and Learning Survey (ETLS) is an EU-wide survey conducted in 2024 to collect representative data on job-related skills development and continuing learning among adult workers in the 27 EU Member States, as well as Iceland and Norway. It gathers information on sociodemographic characteristics, job and firm features, skills development, participation in employer-provided and personal learning activities, attitudes towards learning, workplace conditions affecting skills development, work relationships and labour market outcomes such as wages and job satisfaction.
The ETLS also includes specific questions on informal learning in the workplace. Respondents are asked how often, on their own initiative, they engage in activities to learn new things for their work, such as using books, manuals or audio-visual materials; learning through doing their work; trying new ways of performing tasks; seeking advice from more experienced colleagues or supervisors; observing colleagues at work; or reflecting on ways to improve their performance. Frequency is measured on a five‑point scale ranging from “very often” to “never”. A follow-up question assesses the perceived impact of these activities, asking respondents to what extent they have helped them become better at their job, with response options ranging from “to a large extent” to “not at all”.
Source: CEDEFOP (2026[2]), European Training and Learning Survey, www.cedefop.europa.eu/bg/projects/european-training-and-learning-survey.
The international surveys in this section focus on informal learning by adults. The reason for this is that informal learning among young people has received considerably less focus and the availability of data for its measurement is patchy and fragmented. The OECD’s Programme for International Student Assessment (PISA), provides a small number of proxy indicators of informal learning in some rounds, including youth participation in extra-curricular activities and more direct questions exploring leisure activities, particularly whether young people read, listen or view information materials (e.g. tutorials, podcasts) during their leisure time in order to learn how to do something (OECD, 2021a[4]; OECD, 2021b[5]).
Another example is the European Survey on the use of Information and Communication Technologies (ICT‑HH), which examines internet use among young people, including taking online courses, participating in social networks, and reading news online. Moreover, the European Union Statistics on Income and Living Conditions (EU-SILC) collects data on cultural participation among young people, including visits to museums, art galleries and other cultural sites, while the Harmonised European Time Use Surveys (HETUS) capture time spent visiting libraries (Eurostat, 2019[6]).
Eurostat Adult Education Survey (AES)
The AES, co‑ordinated by Eurostat, has conducted four rounds of data collection to date. The initial 2007 round served as a pilot, followed by subsequent waves in 2011, 2016 and 2022. Each iteration of the survey has included a module on informal learning, enabling the production of comparative statistics on adults’ engagement in this form of learning. The survey defines informal learning as “intentional, but it is less organised and less structured ... and may include for example learning events (activities) that occur in the family, in the workplace, and in the daily life of every person, on a self-directed, family-directed or socially-directed basis”.
The AES is a cross-sectional survey, meaning it captures data at a single point in time and does not follow individuals longitudinally. This limits its utility in assessing the long-term benefits of informal learning. However, it remains valuable for identifying correlations between engagement in informal learning and a range of socio-demographic and socio‑economic characteristics, including age, gender, educational attainment, occupational status, and sector of employment. These variables, highlighted in the literature as antecedents of informal learning, can help uncover patterns of participation. As discussed earlier, the benefits of informal learning may themselves act as further drivers of engagement. While cross-sectional data cannot establish causality, it can provide robust descriptive insights into who participates in informal learning and how these individuals differ from non-participants, for instance in relation to their employment outcomes.
Informal learning in the AES 2022
The Classification of Learning Activities (Eurostat, 2016[7]) conceptualises adult learning through nine attributes mapped onto the triadic classification of learning forms (see also Chapter 3 for a description of the CLA). Within this framework, informal learning is defined as intentional, yet it lacks the other attributes associated with formal and non-formal education and training. While it may occasionally involve predetermined teaching and learning methods, informal learning is primarily characterised by the absence of certain structural features. Specifically, it is not embedded within an organised institutional education and training system, is not linked to a hierarchical qualification structure, does not involve formal admission or registration procedures, is not required to have a specific duration, and its outcomes are not formally recognised by national education or equivalent authorities.
The CLA is employed to operationalise participation in formal, non-formal, and informal learning activities within the AES. As outlined in Chapter 3, the CLA defines informal learning as intentional, a feature made explicit in the AES through the reference to “deliberate learning” in its survey question on informal learning engagement:
‘During the past 12 months, have you deliberately tried to learn anything on a particular topic or area, or are you currently doing it.’
Respondents are then presented with five informal learning activities and can indicate “yes” or “no” for each of them.
(1) Learning from a family member, a friend or a colleague
(2) Learning by using printed material (books, professional magazines, etc.)
(3) Learning by using electronic devices (online or offline)
(4) Learning by guided tours in museums, historical or natural or industrial sites
(5) Learning by visiting learning centres (including libraries).
The five activities used to capture informal learning in the AES encompass opportunities to learn individually or in groups, as well as interaction with non-human learning tools such as books and electronic devices. Although informal learning is conceptually defined by its non-organisational character, the inclusion of “guided tours” appears inconsistent with this definition, particularly in the absence of any reference to self-guided tours. Moreover, while the AES records whether individuals engage in these five types of informal learning activities, it does not include follow-up questions. As a result, it offers limited insight into the depth, purpose, or contextual nature of respondents’ informal learning.
The classification of informal learning activities in the AES also diverges from the categories and classes proposed in the CLA. The CLA distinguishes between “taught learning” – which includes activities such as coaching, informal tuition, and guided visits – and “non-taught learning,” which encompasses self-directed learning, learning groups, practice, and non-guided visits.
Distinguishing informal learning from other types of learning in AES
During the development of the AES survey, countries reported difficulties in clearly distinguishing informal learning from other types of learning. In the evaluation of the 2007 pilot, they specifically highlighted challenges in differentiating informal learning from self-directed study, as well as guided on-the‑job training from training that is an inherent part of the job. Notably, one of the key issues was distinguishing informal learning from non-formal education and training, which in the AES is operationalised as:
‘Activities in which the respondent participated during the last 12 months with the intention to improve knowledge or skills in any area (including hobbies) either during leisure time or working time’.
Respondents are presented with four types of non-formal education and training. They can confirm participation for each of these separately. The four types are presented as follows:
(1) courses at the workplace or in your free time
(2) workshops or seminars at the workplace or in your free time
(3) guided on-the-job training, i.e. planned periods of training, instruction or practical experience, using normal tools of work, either at the immediate place of work or in a work-situation, with the presence of a tutor
(4) private lessons with the aid of a teacher or tutor for whom this is a paid activity.
Informal learning should in particular be differentiated from guided on-the‑job training. This type of learning is informal – rather than non-formal – when it is not an activity which is intentionally planned in advance and is not bound to special or specific places (e.g. classes) or to mediators (e.g. teachers).
Because of these difficulties, Eurostat suggested to entirely delete the module on informal learning after the pilot of 2007. This idea was rejected by the AES Task Force given AES represents a unique source to capture informal learning while Eurostat’s Labour Force Survey’s focus is on organised education and training. The CLA and latest rounds of the AES sought to clarify distinctions by providing concrete examples and clearer definitions.
Changes in definitions and questionnaires in AES
Since the pilot study in 2007, the questions on informal learning changed over the years. Especially the 2011 AES adopted a different approach to measuring informal learning. Unlike the 2007 pilot, where respondents answered “yes” or “no” to a list of activities, the 2011 version began with a general question about whether adults had intentionally tried to learn outside of formal or non-formal education. Those who answered “yes” were then asked to specify the field, the purpose (job-related or personal), and their main method of learning, choosing from four options (i. through a friend, family or colleague; ii. printed materials; iii. Computers; iv. television / radio / video). This reversed the order of questioning, requiring respondents to first self-identify informal learning before providing details. However, due to the complexity of the questions, the 2011 data were deemed not comparable with the pilot, leading Eurostat to return to the original 2007 question format in AES 2016 (Eurostat, 2017[8]). In the meantime, the CLA was revised and updated from CLA2006 to CLA2016.
Since 2016, additional changes have been made to the questionnaire. The question on learning by using electronic devices (online or offline) was introduced in 2022, which merged previous questions on learning by using computers (online or offline) and learning through television/radio/video. This update reflects the increased use of devices such as tablets and smartphones among the adult population. The pilot survey also included a question to identify the main “subject” of learning, but this question is no longer asked in AES2022.
Challenges with the operationalisation of AES in countries
In first rounds of AES, there were some challenges with the comparability of the data on informal learning between countries. Stricter guidelines and support were provided for AES2016 and Eurostat judges that data became more reliable and consistent across countries and that major inconsistencies have been softened by AES2022. However, multiple factors still contribute to possible inconsistencies.
First, the changes in questionnaires do not always appear to be consistently applied in countries. For instance, the new category of learning by using electronic devices (online or offline) – which merged previous categories of learning by using computers (online or offline) and learning through television/radio/video – appears to not have been adopted in all countries. However, in Cyprus, the survey of 2022 still has the category “learning by using computers (online or offline)”, while dropping the category of learning through television/radio/video. In Ireland, they kept the two old categories but slightly rephrased both – i.e. “learning by computers, tables, or smartphones (online or offline)” and “learning through television/radio/videos/DVDs”. These are just two examples, but many countries appear to not strictly follow the guidelines on the phrasing and framing of the questions.
Second, related to the first point is that there is some flexibility in how countries translate and formulate questions. AES works with an international codebook which countries need to translate into their own national language. As such, there can be variation in the formulation of questions. Especially in the AES pilot phase – according to Eurostat’s own quality assessment – countries struggled with providing clear explanations on the differentiation between formal, non-formal and informal learning. Additionally, Eurostat’s quality report for AES2011 states that questions on informal learning were not entirely streamlined across all countries due to flexibility in translation.
However, even in AES 2022, there remains a degree of inconsistency in the translation and formulation of questions related to informal learning. This is evident when comparing the Dutch-language questionnaires used in the Netherlands and Belgium (Flanders) with the official AES manual (see Table 4.1). Although both countries share the same standard language, differences appear in both the wording of the main question and the formulation of response categories.
For example, the manual specifies that the question should ask whether the respondent deliberately tried to learn something on a particular topic or area. However, the Dutch questionnaire omits these terms and instead asks: “Have you, in the past 12 months, tried to learn something yourself by…”, followed by the different categories of informal learning. In contrast, Belgium follows the manual more closely in its phrasing.
There are also several differences in the response categories. The category describing learning through the use of electronic devices (online or offline), as outlined in the manual, is phrased in the Netherlands as: “Using a computer, laptop, tablet, or smartphone?”, with detailed examples in brackets. In Belgium, the equivalent is phrased as: “By using online material (articles, podcasts, apps ...)”. Neither version uses the manual’s terminology of “electronic devices,” opting instead for localised alternatives.
These variations in translation and formulation undoubtedly influence how respondents interpret and answer the questions, thereby likely affecting the level of informal learning reported by these surveys. The fact that such differences exist between neighbouring countries sharing the same official language suggests that disparities in the questionnaires used across other European countries may be even greater. It should be noted that Eurostat provides clear guidelines and standard questionnaires for EU-wide surveys, but it does not have strict control over their implementation in each country.
PIAAC, Survey of Adult Skills 2023
PIAAC provides data from 29 countries collected in the 2023 wave, encompassing 151 711 adults aged 16‑65. PIAAC has been conducted in two full cycles: the first cycle began in 2011 and concluded with its final round in 2019, while the second cycle launched in 2022. Each cycle included multiple rounds of data collection with different countries participating in each round.
The survey includes a direct measurement of skills on literacy, numeracy and adaptive problem-solving. The skills assessment is supplemented by a background questionnaire, which collects data on the participation in formal education and organised training. Similar to AES, it contains additional variables on the typical socio‑economic and socio-demographic background variables such as age, sex, occupational status and educational attainment.
Informal learning in PIAAC
While questionnaires and codebook do not explicitly use the term “informal learning”, the term is mentioned in the “PIAAC Conceptual Framework of the Background Questionnaire Main Survey” document (OECD, 2011[9]), published in November 2011. The following questionnaire items are considered to describe informal learning:
How often does your current job involve learning new things? (H2_D09a)
How often does your current job involve learning-by-doing from the tasks you perform? (H2_D09b)
How often does your current job involve keeping up to date with new products or services?" (H2_D09c)
These indicators are included in the indicator group “learning at work” and the question is therefore only asked when respondents are currently employed. The indicator forms part of the information collected on work tasks in Cycle 2 of PIAAC. In this cycle, the wording of the first question was slightly revised: whereas in Cycle 1 it referred to learning new things “from co-workers or supervisors”, this reference was removed in Cycle 2. The revised phrasing rendered the question more general, potentially broadening its interpretation to include a wider range of informal learning sources in the workplace. As a result, any comparison of informal learning trends between Cycle 1 and Cycle 2 must take this change in wording into account.
PIAAC does not employ binary “yes” or “no” response categories for these questionnaire items. Instead, respondents are asked to choose from a range of frequency-based options: “never”, “less than once a month”, “less than once a week but at least once a month”, “at least once a week but not every day” and “every day”. This approach allows for more nuanced insights into the frequency of informal learning, moving beyond a simply “yes” or “no” dichotomy.
The 2024 Reader Companion for Cycle 2 includes a footnote to explain its conceptualisation of education, training and learning activities. The following note can be found at OECD (2024[10]):
Formal education and training refers to activities that are institutionalised, intentional and planned through public organisations and recognised private bodies. Non-formal education is also institutionalised, intentional and planned by an education provider but leads to qualifications that are not recognised by national educational authorities and can also lead to no qualifications at all. Informal learning takes place outside of institutionalised settings and arises from the learner’s involvement in activities that are not undertaken with a learning purpose in mind.
Additional indicators relevant for informal learning in PIAAC
There are several other indicators in PIAAC that may offer insight into informal learning. The Survey of Adult Skills 2023 Reader’s Companion (OECD, 2024[10]) indicates that informal learning is conceptualised as something that is non-intentional and that generates learning as a by-product of involvement in activities and work tasks, which is distinct from AES which treats informal learning as a deliberate activity. In this context, PIAAC’s information on skills use, both at work and in daily life, and tasks performed at work – such as problem-solving, collaborating with colleagues, etc. – can serve as proxies for informal learning.
PIAAC documentation explicitly cautions that skills use should not be equated with proficiency, and that no information is collected on the complexity of tasks performed. Data on the use of skills in the workplace are intended to shed light on the characteristics of the respondents’ working environment. Additionally, the survey does not capture whether adults perceive themselves as becoming more proficient through learning by doing. Consequently, no claims are made regarding the effectiveness of skills use in generating learning outcomes. For formal and non-formal training, PIAAC does have questions on how useful it is for the respondent’s job, thereby providing some insights on the impact of learning.
PIAAC also includes supplementary questions designed to deepen understanding on workplace characteristics, some of which align with the known drivers of informal learning at the group or organisational level (see respective section in Chapter 2). For example, changes in the workplace may prompt workers to engage in informal learning in order to remain adaptable. Employees engaged in short, repetitive tasks may have fewer opportunities to learn informally than those whose roles require them to continuously respond to new challenges and changing tasks. Factors, such as work pressures and the degree of task discretion or flexibility to organise work autonomously can also influence the time and autonomy available for learning. While the data do not support causal inferences between working environments and the engagement with or outcomes from informal learning, they nonetheless provide valuable contextual information on workplace settings and correlations can also give valuable insights.
Understanding cross-survey variation in reported informal learning participation
Significant differences in participation rates are observed across major international surveys, including the AES and PIAAC. Several factors contribute to these discrepancies. Research in the field of adult learning and education has applied the Total Survey Error (TSE) framework to explain why different surveys produce varying participation estimates (Widany et al., 2019[11]) (see Figure 4.1). Furthermore, data collected within a single survey programme may be affected by measurement or sampling errors, resulting in outcomes that do not fully reflect reality. In the case of informal learning, survey design directly shapes reported participation rates, creating a risk of under- or overestimation (Baldan Babayiğit et al., 2025[12]).
Source: Groves et al. (2004[13]), Survey Methodology.
The TSE framework highlights that survey statistics are shaped by methodological decisions made throughout the survey lifecycle, by underlying methodological choices made throughout the design and implementation of the survey. It recognises that survey results are influenced by a range of possible errors across all stages of the survey process related to both the measurement (e.g. data processing mistakes) and the representation (e.g. coverage and sampling errors). TSE helps researchers identify how methodological decisions affect data quality and comparability.
Variation in conceptualisation of informal learning
Surveys use different conceptualisations of participation, for example on the intentionality of informal learning. The operationalisation of these concepts and the framing of research questions can lead to different interpretations among respondents. Baldan Babayigit et al. (2025[12]) investigated the differences in wording between four survey questionnaires to demonstrate conceptual variation. Questions on participation contain different timespans, ranging from four weeks to learning during the last three years. A shorter time span can fail to capture recent learning while a lengthier one can be distorted by recall issues from the side of the respondent. These issues can lead to validity issues, questioning the accuracy of the intended measurement.
As outlined above, AES and PIAAC are based on distinct conceptualisations of informal learning (see Table 4.2). In AES, informal learning is measured through engagement with human resources (e.g. learning from a family member, a friend or a colleague) or non-human resources (e.g. learning by using electronic devices) and can take place individually or socially. AES defines informal as learning that is intentional and aimed at acquiring new knowledge or skills. In contrast, PIAAC situates informal learning within the context of job tasks, framing it as learning that may occur either incidentally or as a functional response to maintain effectiveness and productivity at work.
Differences in response formats
In addition to conceptual differences, the response formats used in surveys measuring informal learning vary significantly, influencing both the nature of the data collected and its subsequent interpretation. For example, the AES employs a simple binary response format, asking respondents to indicate whether they have participated in informal learning activities by answering “yes” or “no”. While this approach offers a clear measure of participation, it does not capture the frequency, intensity, or duration of engagement.
In contrast, PIAAC collects data on how frequently respondents engage in informal learning, such as learning new skills or knowledge through job-related tasks. This format allows for a more nuanced understanding of learning behaviours, distinguishing between occasional and regular participation.
These differing response formats shape the interpretation of results. The binary format used in AES may yield higher reported participation rates, since even a single instance of informal learning within the reference period qualifies as participation. Conversely, PIAAC’s frequency-based approach can reveal a broader spectrum of engagement highlighting not only whether learning occurs but also how actively individuals participate over time – despite covering a more limited range of informal learning activities. Consequently, comparisons between the two surveys can be challenging, as the data reflect fundamentally different measurement frameworks.
Variation in data collection methods
Variation in data collection methods can also explain differences between surveys, as stipulated by the TSE paradigm. The surveys employ different data collection modes such as CAWI (Computer-Assisted Web Interviewing), CAPI (Computer-Assisted Personal Interviewing), CATI (Computer-Assisted Telephone Interviewing) or PAPI (Paper Assisted Personal Interview) (see Box 4.2). Self-completion formats, such as CAWI, limit opportunities for respondents to seek clarification, potentially affecting response quality. However, well-designed, self-explanatory instructions can help reduce ambiguity (Baldan Babayiğit et al., 2025[12]).
The AES notably employs a combination of data collection modes. In AES 2022, most countries used mixed methods, combining CAPI, CATI and CAWI, while a limited number used PAPI or other alternatives. For example, in the Czech Republic (hereafter “Czechia”), 49% of respondents were interviewed via PAPI. In Belgium, 76% of respondents completed the survey through CAWI and 24% through CAPI. The choice of mode was largely driven by cost considerations and the diversity of the adult population.
Some of the most common data collection modes are:
CAWI (Computer-Assisted Web Interviewing): Respondents complete the survey online via a web interface, typically on their own device. This method is cost-effective and allows for flexible scheduling but may exclude individuals with limited internet access or digital skills.
CAPI (Computer-Assisted Personal Interviewing): An interviewer conducts the survey face‑to-face using a computer or tablet to record responses. This method allows for more complex surveys and clarifications but is more costly and time‑consuming.
CATI (Computer-Assisted Telephone Interviewing): An interviewer administers the survey by phone and enters responses into a computer. This method offers a balance between reach and cost, though it may limit the use of visual materials and is subject to declining response rates due to call screening.
PAPI (Paper Assisted Personal Interview): Face‑to-face interviews where responses are recorded on paper forms.
While most countries used a mix of data collection methods, a few relied on a single mode. For instance, the Netherlands used only CAWI to collect responses. Similarly, Bulgaria, Greece, Romania, and Serbia each employed just one mode. Luxembourg primarily used CAWI but collected 3.7% of responses via offline methods. A majority of countries applied two modes – typically combinations of CAPI, CATI or CAWI – while Spain, Lithuania and Latvia employed all three.
Survey mode influences completion times, with notable differences observed across modes. For example, in Belgium, the average completion time for the AES was 12 minutes for respondents using CAPI, compared to 22 minutes for those completing the survey via CAWI. These discrepancies are important to consider, as they may affect data quality, contribute to respondent fatigue, and ultimately impact the reliability of survey results.
Since differences in participation rates are influenced by a wide range of factors (see also overview of drivers in Chapter 2), it is difficult to determine whether variations in national data collection methods had a significant impact on the results. Explaining the differences in AES participation rates for 2022 would require additional analysis of the microdata, controlling for a broad set of factors.
In comparison to AES, PIAAC used a streamlined CAPI method to collect its data. Interviewers collected data on tablets (OECD, 2024[10]). As part of their work with respondents, interviewers were also asked to log some of their observations. For example, they were asked to include their judgement on whether the respondent generally understood the survey questions.
Differences in survey populations
Statistics on participation in adult learning and education also depend on the target population of the study, in line with the possibilities of representation errors as stipulated by the TSE. European benchmarks on “education and training for adults” target those between the ages of 25 and 64 although surveys include data on a wider range of younger and older adults. Sampling frameworks are guided by various population registers or Census data dependent on the location and differ in their use randomisation and stratification methods. Survey datasets typically come with population weights to correct for respondent selection and non-response with the aim to represent the target population more accurately. However, calibration approaches also differ across surveys, meaning that the representativeness in one dataset may not align with that in another.
Several countries require selected citizens to participate in Eurostat surveys. For the AES 2022, participation was mandatory in Spain, France, Croatia, Italy, Cyprus, Luxembourg, Malta, Portugal, the Slovak Republic, and the Republic of Türkiye (hereafter “Türkiye”) (Eurostat, 2025[14]). In some countries, voluntary participation is associated with high unit non-response rates, the highest being in Belgium (81.5%). Other countries where more than two‑thirds of selected adults did not respond include Denmark (73%), Germany (69.9%), and the Netherlands (69.8%). Notably, Italy recorded a non-response rate of 70% despite mandatory participation. The lowest non-response rate among countries with voluntary participation was in the Czechia (12.7%). People who are not responding to surveys, do so for various reasons, including for instance not having the time, not having the required language skills, and more.
In the case of AES and PIAAC, differing definitions of informal learning also shape the relevant target populations on this specific topic. AES administers its informal learning module to all respondents, whereas in PIAAC, the questions related to informal learning – specifically those in section H2, “learning at work” – are posed only to employed adults. As such, rates on informal learning derived from the PIAAC dataset can only be generalised toward the working population.
The challenge of self-reported measures for informal learning
A common concern with both AES and PIAAC is their reliance on self-reported measures. Respondents must recall and recognise the examples of informal learning provided in the questionnaire (Baldan Babayiğit et al., 2025[12]). In AES, instructions focus on deliberate informal learning, for example through interactions with colleagues. In PIAAC, respondents must first be aware of their own learning before acknowledging it. Such measures therefore depend on how respondents interpret the questions, rather than on objective, fact-based statistics transferable from registration data (Robinson and Leonard, 2024[15]). Robinson and Leonard (2024[15]) argue that survey questions should be easy to understand and elicit accurate interpretations of the data collectors’ intent.
In practice, survey respondents often face challenges in understanding questions. This is illustrated by PIAAC, in which interviewers record the extent to which respondents understand the questions. While this is not specific to items on informal learning, the data indicate that across the OECD, nine in ten adults often or very often understood the questions. However, according to the interviewer, 2.4% never or almost never understood them, and a further 7.8% understood them only “now and then” (see Figure 4.2). Differences across countries are considerable: Japan had the highest overall share of adults who did not often or very often understand the questions (29% of respondents), while the Flemish Region (Belgium) had the largest share of adults who never or almost never understood them (6.5%). Overall, these findings illustrate the limits of self-reported measures and the potential for misinterpretation.
The use of self-reported measures in AES and PIAAC risks underestimating informal learning among groups less familiar with the terminology used in the questions. Programmes continue to rely on self-reporting, largely because it is inexpensive and because objective data are often unavailable. While registration data on formal education may become increasingly accessible for research in some countries (see e.g. Donnelly et al (2024[16]); Mellander (2017[17])), the absence of a dedicated organisational framework makes it difficult to capture comparable, objective data on informal learning. Advances in alternative data collection and analytical methods may in time allow for a more detailed study of such learning patterns beyond reliance on traditional survey approaches. A related challenge is the varying degrees of self-selection, linked to substantial non-response rates that differ across countries, as described above.
Source: OECD (2025[18]), OECD Survey of Adult Skills (PIAAC), www.oecd.org/skills/piaac/.
National approaches to the measurement of informal learning
International surveys – as described above – provide vital measures of informal learning that are comparable across countries, and they are widely used by policymakers and wider stakeholders. However, the breadth and depth of the measures included in international survey instruments are limited, often reflecting space constraints in survey questionnaires, the trade‑off between survey length and response rates and a general need for international organisations to prioritise questions of greatest relevance to policymakers across the EU and OECD.
National survey instruments can therefore complement international surveys in a range of ways. First, they can be conducted more frequently, enabling policymakers to track trends in participation and consider the efficacy of policy measures designed to promote informal learning. Second, they can accommodate a more extensive set of questions examining motivations for learning, types of activities, skills acquired, outcomes and benefits. There is scope for expanded national samples that enable more in-depth interrogation of the data and examination of the relationships between different variables; as well as longitudinal studies that track cohorts over time. Third, while international surveys ensure comparability across contexts and alignment with global standards, national surveys offer the flexibility to address country-specific objectives, such as employer participation, workforce development, or RPL.
A survey conducted for this project offers important insights into how informal learning is measured within national education and skills systems. Responses from experts in 15 OECD countries reveal a range of approaches, with most relying on nationally administered surveys as the primary source of information.
In addition, 11 of the 15 responding countries reported measuring informal learning primarily though international survey frameworks, most notably the AES and PIAAC (see Figure 4.3). These internationally co‑ordinated instruments are typically adapted at the national level and serve as a common basis for data collection on learning activities beyond formal and non-formal education. For instance, countries such as the Czechia, Latvia, and Italy report relying on such international surveys to monitor and analyse participation in informal learning, helping ensure data comparability across countries.
Responses to questions “Is informal learning measured in your country” (N=15)
Source: OECD analysis and a policy questionnaire completed by countries.
In 4 out of 15 responding countries, informal learning is measured through independent national surveys or data collection initiatives designed to reflect the learning experiences of the adult population. These instruments are typically tailored to national contexts and aligned with specific policy priorities. For example, in the Netherlands, the Monitor Leercultuur (Learning Culture Monitor), led by the Social and Economic Council (SER) and the Netherlands Organisation for Applied Scientific Research (TNO), tracks trends and changes in workplace learning behaviours (SER, 2024[19]). It places particular emphasis on informal learning, including collaboration, task novelty, and feedback from colleagues, supervisors, or customers. Furthermore, the Netherlands conducts the Lifelong Learning and Development Survey (ROA, 2024[20]), carried out every three years since 2004. This survey includes a quantitative measure of informal learning, based on the question: “What percentage of your working time do you spend on tasks from which you can learn?” Besides providing a trend of informal learning over time, it also allows comparison with time spent on non-formal education (De Grip, 2024[21]).
Similarly, in Australia, the Employers’ Use and Views of the VET System Survey (SEUV), conducted biennially by the National Centre for Vocational Education Research (NCVER), provides insights into employer engagement with the VET system. The SEUV includes a dedicated category for informal training, defined as unstructured, non-credentialled, and plan-free learning. Examples include on-the‑job learning, self-directed study using manuals or software, mentoring, and peer learning. In the United Kingdom, the Skills and Employment Survey (SES) also features a wide range of questions exploring participation in various informal learning activities, motivations for engagement, associated outcomes, skills acquired and the application of these in the workplace (Butt et al., 2025[22]). Finally, in Italy, the National Institute for the Analysis of Public Policies (INAPP) conducts the Survey on Adult Learning Behaviours (INDACO-Adults), which includes questions on participation in intentional informal learning activities harmonised with the AES, as well as additional questions exploring the frequency with which respondents engage with wider unintentional informal learning activities and their perspectives on “micro-learning” as an area of national policy interest.
These examples reflect a growing interest among countries in understanding informal learning as a core element of lifelong learning strategies. While international surveys ensure comparability across contexts and alignment with global standards, national surveys offer the flexibility to address country-specific objectives, such as employer participation, workforce development, or the recognition of prior learning.
However, the examples of surveys also highlight the significant differences between survey instruments that can hinder international comparability. Definitions differ in scope and operationalisation, with some focussing on intentional informal learning and others capturing incidental learning, often without explicitly referencing underlying conceptual frameworks. Survey design also differs in terms of target populations (employers versus individuals), age cohorts, modes of data collection and methodological choices such as sampling and weighting, affecting comparability. Question wording and response formats show particularly wide variation, ranging from objective measures of participation over defined periods to subjective assessments of learning opportunities and workplace conditions, using binary, scale‑based or quantitative response options. Interpretation and use of data also differ, from standalone research reports to integrated monitoring systems such as national lifelong learning “monitors”. Overall, while efforts to strengthen measurement are ongoing, they remain fragmented, with limited cross-country collaboration among researchers working on comparable datasets.
Alternative approaches to measuring informal learning
While national and international survey instruments provide vital means to measure, compare and track trends in informal learning, informal learning can also be assessed through a range of alternative methodologies. Academics and researchers continue to experiment with innovative approaches to measure the scale, patterns, conditions, drivers, outcomes and benefits of informal learning. The concept of moving beyond the traditional triadic classification of learning modes is an example of such an alternative approach (see Box 4.3).
A recent publication proposes two alternative scenarios to measure participation in adult learning, moving away from the triadic classification. The fuzzy-set approach helps to identify overlaps between the nine attributes of the CLA, revealing whether they truly belong to formal, non-formal or informal categories, or whether they operate as hybrid modes between them. The learning continuum scenario borrows insights from workplace learning to measure the characteristics of informal learning through Likert scales, leading to an overall score on a scale that captures the full continuum from informal learning to formal education and training.
Source: (Kalenda and Boeren, 2025[23]), Revisiting the triadic classification of learning activities: Rethinking their measurement, International Journal of Lifelong Education, Vol. 1‑16. https://doi.org/10.1080/02601370.2025.2489506.
Other examples of alternative approaches for the measurement of informal learning are:
Time‑use or diaries studies: These studies are designed to capture informal learning as part of daily life. They record what people do, when, for how long, and with whom, sometimes adding short reflections on purpose or outcomes. Time‑diary designs are particularly helpful in identifying learning opportunities embedded in routines, which often constitute unintentional, and sometimes unconscious, informal learning. They can also provide insights into the frequency or intensity of learning, which are often challenging to quantify through standard surveys. Researchers in Ireland, for example, have used diary studies to examine reading for pleasure and engagement in cultural activities among 9‑year‑olds, utilising data from the Growing Up in Ireland study (Smyth, 2022[24]).
Ethnographic studies: These qualitative studies can also be helpful in capturing both intentional and unintentional informal learning, observing informal learning as it happens, embedded in work, play and daily life rather than relying on recall-based survey questions. Recent ethnographies of workplace learning, for example, combine in situ observation with interviews to document how people learn through improvisation, asking colleagues, copying practices, and feedback – important forms of informal learning at work through social interaction. There is also growing interest in digital ethnography, examining digital practices that promote informal learning (Karhapää, Hämäläinen and Pöysä-Tarhonen, 2025[25]).
Social network and interaction analysis: These approaches play an important role in measuring informal learning through social interaction. Methods range from surveys exploring where individuals seek advice or support to trace‑based interaction networks (e.g. via social networks or digital platforms) and the use of wearable sensor technology to study the dynamics of social learning in the workplace (Endedijk and Van den Bossche, 2017[26]).
Learning analytics: These approaches can make informal learning patterns visible where participation leaves digital traces (views, posts, replies, contribution histories, navigation sequences). Analysis of this data offers the potential to study patterns of informal learning on an individual or social basis through the use of online learning or digital platforms, for example.
Examples of these methodologies are presented in Table 4.3, showing how informal learning can be captured across different populations, settings, and time frames using complementary approaches. Together, they highlight that informal learning is frequent, embedded in everyday practices, socially mediated, and increasingly digital, and that combining methods can better capture its context, intensity, and dynamics than recall-based surveys alone. The insights from these innovative methodological approaches can complement survey evidence and help inform the design of national and international instruments.
This section describes the third priority topic on “Strengthening international survey instruments”. To strengthen international survey instruments, the OECD recommends the following actions: i) preserve and extend the measures of informal learning included in international survey instruments; ii) promote greater consistency in the implementation of international surveys at the national level.
Recommendation 5: Preserve and extend the measures of informal learning included in international survey instruments
Considerable progress has been made in operationalising the concept of informal learning in international survey instruments, and interviews with policymakers and international experts suggest they play a vital role in enabling countries to benchmark performance against their peers and track progress over time. Indeed, of the 15 OECD countries that responded to the survey conducted for this project, 11 reported measuring informal learning primarily through international surveys, such as the AES and the PIAAC. It is therefore vital that indicators of informal learning in future waves of the AES and PIAAC are retained, as vital mechanisms for benchmarking performance and tracking progress towards policy aspirations to stimulate informal learning (see Recommendation 2).
Further, the OECD recommends that data owners explore the scope to strengthen the measurement of informal learning in international surveys, in a range of ways.
Refine measures to capture intensity: it is recommended that international survey instruments move beyond binary measures of participation (“yes” or “no” responses) and adopt indicators capturing the intensity of informal learning. Given the broad conceptualisation of informal learning, participation rates can be very high – up to 93% in some OECD countries, according to the AES (Eurostat, 2022[34]) – and emerging forms, such as learning via social media, interactive games or AI tools, are likely to increase this further and also raise questions about the depth of learning. Existing binary measures offer limited insight into cross-country differences, trends over time, or the extent of skills and knowledge acquired.
Data owners are encouraged to reframe questions to capture the frequency of engagement as a proxy for intensity. For example, response scales that are already employed in PIAAC could be used (e.g. ”never”, “less than once a month”, ‘less than once a week, “at least once a week” and “every day”).
Capture the benefits and outcomes of learning: there is scope to include additional questions exploring the purpose of informal learning and the perceived outcomes and benefits of participation. For instance, these could be questions exploring the extent and nature of knowledge or skills acquisition associated with participation in informal learning activities.
Potentially, the questions could be based on ones from existing international surveys, such as those in the AES on outcomes of formal and non-formal education and training – e.g. whether acquired skills or knowledge are applied and whether learning contributed to a new job, promotion, new tasks, improved performance, or other relevant benefits.
Capture the conditions of informal learning: There is growing interest in the conditions that support informal learning in various settings, including in schools and at work (e.g. specific workplace practices or workplace cultures). While data owners need to avoid making surveys overly long and complex, which may affect response rates, the inclusion of such questions would enable policymakers to evaluate the impact of interventions that seek to strengthen the learning culture within enterprises or promote innovative pedagogical approaches. Additionally, questions that explore the conditions for informal learning can also better support the measurement of unintentional informal learning, where self-reported measures are less suitable.
Extend the measurement of informal learning across the full typology: Given the breadth of its conceptualisation, it is unlikely that a single data source will be sufficient to capture informal learning in its varying forms. Instead, there is a need to consider how different international surveys might work together to provide a comprehensive and coherent international framework for the measurement of informal learning. This could include mapping sources and indicators across the new typology of informal learning, to identify areas of overlap and evidence gaps. Figure 4.5 provides an illustrative mapping, highlighting potential to widen measures in schools to include informal learning through social interaction, while pointing to scope to rationalise and align varying measures of informal learning in the workplace, for example.
Continuously update surveys for emerging learning forms: There is a need to continue to review survey questionnaires to ensure they adequately capture emerging forms of informal learning (e.g. the use of AI tools). Wherever possible, this should entail updating response options or guidance notes, rather than introducing new questions that might introduce discontinuity between survey waves. This mirrors the approach taken by Eurostat in updating the AES informal learning questions to reflect increased use of digital devices, such as tablets and smartphones.
Note: The colours of the survey titles reflect the predominant learning setting captured by each survey (e.g. purple indicates school, blue indicates the workplace, etc.).
Recommendation 6: Promote greater consistency in the implementation of international surveys at the national level
Strengthening the design of international survey instruments will be vital to advance the measurement of informal learning, but an additional challenge lies in the operationalisation of these instruments at the national level.
All international surveys go through a process of national adaptation, where master or model questionnaires are adjusted to different national contexts (e.g. reflecting national qualifications, currencies, etc.) and translated into the national languages of participating countries. International statistics agencies such as Eurostat provide detailed guidance to aid this process, often in the form of manuals, translation and adaptation guidelines, and international codebooks.
While offering flexibility in the formulation and translation of survey questions to allow for cultural differences is important, it is vital that this process does not introduce measurement variance. National Statistics Offices should follow guidance materials closely, maintaining harmonised definitions of informal learning, consistent question formulation and response options at the national level and ensuring that questions are understood similarly by respondents in different countries so that responses can be validly compared and international comparability maintained. To support National Statistics Offices in the consistent implementation of international surveys, several actions can be taken.
First, the complexity of the concept of informal learning means that additional guidance, resources and training for interviewers will be needed. Eurostat has progressively refined the introduction to the section on informal learning in the AES, including clearer framing, a more detailed definition of informal learning grounded in the CLA; and more extensive implementation guidelines that explore different forms of informal learning and using easy-to‑understand examples to build comprehension among interviewers and respondents. However, additional training and resources for interviewers, utilising the harmonised conceptual definition and the new typology of informal learning (Recommendations 1 and 2), could further strengthen comprehension of key concepts and provide practical examples that support interviewers in building understanding among survey respondents.
Second, experts engaged through the project emphasised the importance of the sequencing of questions, for example progressing from formal to non-formal education and training and subsequently informal learning as a means for building understanding of the concept; and contextualising the questions by specifying the setting within which informal learning is being examined (e.g. the workplace).
Third, data owners should explore the scope to utilise new AI tools to assist interviewers and respondents in categorising varying forms of learning. Such tools could prove a valuable addition to existing resources and training that seek to clarify complex concepts, distinguish between different forms of learning and categorise types of informal learning, utilising the new typology and decision trees described in Priority Topic 2. Conversational agents are in growing use in a range of scenarios, from customer service chatbots to clinical triage tools, blending Natural Language Processing (NLP) models with AI to guide users through decision trees. In education and labour market contexts, AI-assisted tools are frequently employed to classify free‑text job titles and descriptions (see Box 4.4), while adaptive learning platforms use AI to assess learner progress and personalise learning pathways dynamically.
CASCOT (Computer Assisted Structured Coding Tool), developed in the late 1990s by the Institute for Employment Research at the University of Warwick, is widely used in the United Kingdom to code open‑ended survey responses into SOC and SIC classifications. Originally based on probabilistic text‑matching, it now incorporates more advanced NLP and machine‑learning techniques to improve semantic matching and ranking. CASCOT suggests likely codes with confidence scores, helping researchers prioritise manual checks and maintain consistency at scale. It is used across major labour‑market surveys, employer skills studies, and SOC‑based administrative processes, reducing manual workload while strengthening the reliability of occupational and industrial statistics.
Source: (Institute for Employment Research, 2026[35]) Computer Assisted Structured Coding Tool, https://warwick.ac.uk/fac/soc/ier/data_group/cascot/.
This section describes the fourth priority topic on “complementing international instruments with national measures”. To complement international instruments with national measures, the OECD recommends the following actions: i) include measures of informal learning in national survey instruments; ii) establish an international Informal Learning Expert Working Group.
Recommendation 7: Include measures of informal learning in national survey instruments
Countries should introduce or expand questions relating to informal learning in national survey instruments. While the project has identified a range of examples of national instruments that measure informal learning (some of which are referenced above), it has also found national learning surveys where questions are limited to formal and non-formal education and training, and surveys exploring working conditions where questions concern only organised training or more structured development opportunities.
Measuring informal learning through national survey instruments offers several key benefits:
Regular monitoring: National surveys are often updated on an annual or biennial basis. Incorporating questions on informal learning into these existing survey instruments would greatly strengthen countries’ ability to monitor participation in informal learning, and at more regular intervals than afforded by international datasets like AES or PIAAC.
Capturing depth and breadth of informal learning: Beyond participation in informal learning, countries should look to utilise national survey instruments to address evidence gaps in international surveys (as highlighted, for instance, by Figure 4.5), extending measurement across the typology of informal learning, and offering additional depth or granularity. This could include questions on:
informal learning in schools
unintentional informal learning though social interaction in various settings
the motivation for informal learning
participation in different types of activities
the intensity of learning
the extent and type of knowledge and skills acquired
the outcomes and benefits realised through informal learning activities.
Analysing variation across populations and contexts: National survey instruments should also be utilised to explore variation across different socio-demographic groups or industry sectors, as well as the drivers behind informal learning. Indeed, stakeholders consulted during this project expressed keen interest in further understanding the conditions that stimulate informal learning in various settings, including but not limited to the workplace.
Supporting evidence‑based policy: Data from national surveys can also support evidence‑based policymaking by including questions related to national policy priorities for informal learning. This will provide baselines and track progress over time, informing the evaluation and refinement of interventions and policies aimed at strengthening informal learning.
It is possible, however, that a single national survey instrument will be insufficient to address all these needs. Instead, countries should consider how multiple survey instruments might work together, each with clearly defined and complementary purposes. Such an approach is evident in the Netherlands, which benefits from a series of national survey instruments, with some designed to regularly track indicators over time, while others provide more in-depth information or longitudinal data to support the assessment of outcomes and benefits (see Box 4.5).
In the Netherlands, there has been a growing focus on informal learning in recent years, as part of wider efforts to strengthen lifelong learning and nurture a learning culture. To provide better evidence to inform policy development, stakeholders have developed national survey instruments to support the measurement of informal learning, alongside international surveys, such as the AES. These national survey instruments are designed to play distinct and complementary functions:
TNO/SER Monitor Learning Culture: The Netherlands Organisation for Applied Scientific Research (TNO) in collaboration with the Social and Economic Council (SER) publishes the Monitor Learning Culture every two years. The latest edition was published in November 2025, and it shows how employed people in the Netherlands learn and how this has changed over time, capturing formal and non-formal education and training, and informal workplace learning. Indicators are grouped into three categories: actual learning behaviour, the perceived need to update skills or match them to work, and supportive factors such as working conditions and HR policies. Data are drawn from several large‑scale national surveys to monitor working conditions of employees, self-employed workers and employers, conducted by TNO and Statistics Netherlands (CBS). Additionally, cohort studies are used to track fixed groups of workers over time. Alongside the Monitor Learning Culture, TNO also publishes in-depth research on specific topics, including a recent study on the factors that influence informal learning and innovative employee behaviour.
ROA Lifelong Learning Survey and the LLO Radar: The Research Centre for Education and the Labour Market (ROA) has conducted its Lifelong Learning Survey (LLO) every three years since 2004. It consists of a core set of survey questions, unchanged between waves, coupled with additional questions that address specific issues and allow for a deeper understanding of participation in different types of informal learning activities, how patterns of participation vary depending on individual and workplace characteristics, knowledge development, skills use and the outcomes of informal learning. More recently, as part of a multi-year LLO-Catalyst (LLO=katalysator) initiative, stakeholders have also sought to bring together data from the Lifelong Learning Survey with the Dutch Skills Survey (NSS) and wider sources (e.g. Labour Force Survey [EBB], PIAAC) to create the LLO Radar, a dynamic system for monitoring trends, bottlenecks and opportunities in the labour market.
Source: SER (2024[36]), Monitor Leercultuur 2023, https://publicaties.ser.nl/ser_leercultuurmonitor_2023/cover; Kunn, Baumann, Fouarge, Hendrickx, and Lansink (2024[37]), Learning and development in the Netherlands, https://cris.maastrichtuniversity.nl/en/publications/leren-en-ontwikkelen-in-nederland/; Kunn-Nelen, Fouarge, Weel and Bussink (2025[38]), Exploring datasets for an LLO Radar, https://cris.maastrichtuniversity.nl/en/publications/verkenning-datasets-voor-een-llo-radar/; Koopmans, van de Ven, van der Torre and de Geit (2025[39]), Informal learning and innovative employee behaviour.
Recommendation 8: Establish an international Informal Learning Expert Working Group
In addition to working independently to strengthen national measures of informal learning, it is advised to establish an international Expert Group on the Measurement of Informal Learning (EGMIL) to promote collaboration between OECD countries and improve cross-national comparability of survey instruments. This group would bring together statisticians, survey methodologists, and researchers from countries actively involved in defining, measuring, and analysing informal learning across different contexts (workplace, home, community, and digital environments).
Building on successful models of international statistical collaboration, the group should provide a structured but flexible forum for technical exchange. It could meet twice a year, complemented by virtual working sessions, to share national practices, discuss conceptual and methodological challenges, and jointly address gaps in existing survey instruments. A key objective would be to clarify and operationalise shared definitions of informal learning, while recognising national diversity in policy priorities and data infrastructures.
The working group could collaboratively develop practical outputs, such as:
examples of informal learning in its varying forms, aligned to the new typology
guidance on question wording and survey design for informal learning
documentation of proxy indicators and their limitations
a repository of validated national survey questions.
Over time, this would promote greater consistency, transparency, and credibility in national measures of informal learning, while reducing duplication of effort and supporting cumulative learning across countries.
Moreover, the international Informal Learning Expert Working Group could play a key role in disseminating knowledge, raising awareness of informal learning, and promoting the adoption of relevant tools and methodologies across countries.
The proposed group could seek to emulate the governance and working methods of existing, comparable groups of experts. Although broader in scope, the UN’s Expert Group on Well‑Being Measurement could provide one such example: a time‑bound, expert-led forum working to advance shared statistical frameworks and deliver concrete methodological outcomes through regular collaboration (see Box 4.6).
The Expert Group on Well-Being Measurement (EGWM) was established in 2024 by the UN Statistical Commission (UNSC) at the request of the UN Network of Economic Statisticians to support global efforts on measuring progress beyond GDP and advancing well-being statistics. The aim of EGWM is to develop the Framework for Inclusive and Sustainable Wellbeing (FISW) – a comprehensive conceptual and statistical framework that combines key well-being dimensions into a coherent set of indicators, including a headline dashboard to inform policymaking and public understanding. Other tasks include drafting framework chapters, proposing FISW text to the UNSC, commissioning pilot compilations of data, and shaping communications strategies.
Membership spans National Statistical Offices (NSO) from countries such as Canada, Italy, the United Kingdom, and the United States, major international organisations (e.g. OECD, IMF, UNDP, WHO), and academic partners like Cambridge University and Imperial College London. The EGWM meets regularly, through formal sessions and additional meetings focussed on indicators and methodology.
Source: The UN Network of Economic Statisticians (2024[40]), Bureau Briefing Paper: Implementing the Expert Group on Well-being Measurement, https://unstats.un.org/unsd/statcom/groups/EGWM.
This section describes the fifth, and last, priority topic on “promoting ongoing innovation in the definition and measurement of informal learning”. To promote ongoing innovation in the definition and measurement of informal learning, the OECD recommends the following actions: i) identify informal learning measurement as an explicit research priority; ii) regularly review and update national and international measurement instruments.
Recommendation 9: Identify informal learning measurement as an explicit research priority
National and international research councils should play a stronger and more sustained role in supporting research that advances innovative and robust methodologies for understanding, measuring and validating informal learning.
In many countries, some of the most promising methodological advances in this area have been driven by government- and Research Council-funded programmes, particularly those supporting interdisciplinary and international collaboration. One example of this is the EU’s Horizon Europe programme that has embedded lifelong learning as a cross-cutting priority across multiple research clusters, supporting multinational consortia to develop new conceptual frameworks, comparative methods and data sources for understanding learning across the life course (see Box 4.7).
However, research activity remains fragmented and often time‑limited. There is therefore a need for ongoing, dedicated funding streams that explicitly support methodological innovation related to informal learning. This should include funding for the development, testing and validation of new indicators and measurement approaches, as well as comparative and cross-national research that improves conceptual clarity and consistency across contexts.
In addition to academically oriented research, research councils could also support the establishment of applied, real-world “testbeds”, such as living labs, workplace‑embedded research environments, and community-based pilots, in settings where informal learning naturally occurs. This could include workplaces, communities, cultural and creative contexts, digital platforms and voluntary or civic organisations. By embedding research within everyday learning environments, such testbeds would enable the iterative testing and comparison of alternative methodological approaches, support co-creation with practitioners, employers and learners, and allow for robust evaluation of how informal learning can be identified, measured and validated in practice. This approach could also facilitate the development of practical tools, guidance and resources that could inform policy development and practical action for stimulating informal learning in different contexts (e.g. in the workplace).
Horizon Europe is the European Union’s flagship research and innovation framework programme for 2021‑2027, with a budget of around EUR 95 billion. It is explicitly designed to promote international, interdisciplinary collaboration on shared societal and economic priorities, requiring most projects to be delivered by multi-country consortia spanning universities, research institutes, policymakers, businesses and civil society.
Lifelong learning (LLL) is a cross-cutting theme, particularly relevant within Cluster 2: Culture, Creativity and Inclusive Society; and Cluster 4: Digital, Industry and Space. Calls under these clusters address adult skills development, reskilling and upskilling, learning across the life course, workplace and informal learning, digital skills, and social inclusion. Many projects explicitly examine how individuals acquire, update and use skills beyond initial education, and how learning systems can better respond to technological change, demographic ageing and inequality.
Horizon Europe also promotes methodological innovation relevant to LLL research, including comparative cross-national designs, mixed-methods approaches, participatory and co-creation methods, and the development of new indicators and data sources on skills and learning. By combining scale, sustained funding and mandatory international collaboration, Horizon Europe functions as a powerful model for advancing shared understanding, measurement and policy-relevant evidence on lifelong learning across countries.
Source: European Commission (2021[41]), Horizon Europe – Framework Programme for Research and Innovation (2021‑2027), https://research-and-innovation.ec.europa.eu/funding/funding-opportunities/funding-programmes-and-open-calls/horizon-europe_en; European Commission (2023[42]), Horizon Europe Programme Guide, https://ec.europa.eu/info/funding-tenders/opportunities/docs/2021-2027/horizon/guidance/programme-guide_horizon_v4.0_en.pdf.
Recommendation 10: Regularly review and update national and international measurement instruments
Digital technologies, new forms of work, and changing patterns of knowledge exchange continue to blur established boundaries between different types of learning and are reshaping how people learn informally. As a result, informal learning should be understood as a dynamic concept.
National policymakers and international organisations should therefore review the definition and measurement of informal learning on an ongoing basis to ensure that they reflect emerging forms of learning and to explore how alternative methodologies could complement or, where appropriate, replace existing surveys. Regular reviews of the wider research literature would help keep pace with methodological advances and support a structured, evidence‑informed approach to updating definitions and measurement practices. At the same time, any changes should be carefully balanced against the need to maintain comparability over time, as frequent revisions to definitions or measurement approaches can make it more difficult to track trends and assess progress.
However, methodological progress in this area cannot be achieved through isolated projects alone. There is a clear need to strengthen knowledge exchange, co‑ordination, and collaboration across the international research community, bringing together expertise from education research, labour economics, sociology, data science, and related fields. The International Expert Group on the Measurement of Informal Learning (EGMIL) (see Recommendation 8) provides a strong starting point for this work.
International research conferences and policy forums offer another important mechanism for supporting exchange, and national policymakers and international organisations could make more systematic use of these platforms. Established and authoritative platforms such as the European Conference on Educational Research and OECD-convened policy forums (see Box 4.8) regularly bring together researchers, policymakers and practitioners from multiple countries. These forums create space to share emerging evidence, compare methodological approaches, and discuss the strengths and limitations of different ways of capturing informal learning across contexts. They also provide opportunities to showcase innovative methods, including applied testbeds, mixed-methods designs and new data sources, and to reflect collectively on their policy relevance.
The European Conference on Educational Research (ECER) is a major international forum for education research, convened annually since 1992 by the European Educational Research Association (EERA). Hosted each year by a European university or research institute, ECER aims to advance high-quality, comparative and policy-relevant research across all areas of education. Lifelong learning is a core focus, addressed through dedicated networks on adult and continuing education and through cross-cutting themes on skills, equity, social inclusion and learning across the life course.
Source: European Educational Research Association (2026[43]), European Conference on Educational Research 2026, https://eera-ecer.de/conferences/ecer-2026-tampere.
References
[12] Baldan Babayiğit, B. et al. (2025), “Unravelling the data puzzle: a total survey error perspective on adult learning and education participation in the UK”, International Journal of Lifelong Education, Vol. 44/3, pp. 273-290, https://doi.org/10.1080/02601370.2025.2493798.
[22] Butt, S. et al. (2025), Skills and Employment Survey 2024: Technical Report, Cardiff University / NatCen, https://wiserd.ac.uk/wp-content/uploads/SES-2024-Technical-report_Amended_040425.pdf.
[2] CEDEFOP (2026), European Training and Learning Survey, https://www.cedefop.europa.eu/bg/projects/european-training-and-learning-survey (accessed on 4 March 2026).
[28] Compagnoni, M. et al. (2024), “Everyday, every week, all at once? An experience sampling study on teachers’ professional development for the classroom, team, and school”, Teaching and Teacher Education, Vol. 152, p. 104771, https://doi.org/10.1016/j.tate.2024.104771.
[21] De Grip, A. (2024), “The importance of informal learning at work - On-the-job learning is more important for workers’ human capital development than formal training”, IZA World of Labor, https://doi.org/10.15185/izawol.162.v2.
[16] Donnelly, M. et al. (2024), “Education and the spatial division of labour: further education and prospects for ‘ <i>Levelling Up’</i>”, Contemporary Social Science, Vol. 19/4, pp. 514-530, https://doi.org/10.1080/21582041.2024.2418119.
[26] Endedijk, M. and P. Van den Bossche (2017), Future perspectives and challenges of using wearable sensor technology to study dynamics of social learning at the workplace.
[32] Endedijk, M. and P. van den Bossche (2017), “Sociometric badges and WiFi sensors: using wearable sensor technology to study dynamics of social learning at the workplace”, Paper presented at 17th Biennial Conference of the European Association for Research in Learning and Instruction (EARLI) 2017, Tampere, Finland..
[1] Eurofound / Cedefop (2019), European Company Survey 2019, Eurofound, https://www.eurofound.europa.eu/en/surveys-and-data/surveys/european-company-survey/ecs-2019.
[42] European Commission (2023), Horizon Europe Programme Guide, European Commission, https://ec.europa.eu/info/funding-tenders/opportunities/docs/2021-2027/horizon/guidance/programme-guide_horizon_v4.0_en.pdf.
[41] European Commission (2021), Horizon Europe – Framework Programme for Research and Innovation (2021–2027), European Commission, https://research-and-innovation.ec.europa.eu/funding/funding-opportunities/funding-programmes-and-open-calls/horizon-europe_en.
[43] European Educational Research Association (2026), European Conference on Educational Research 2026, https://eera-ecer.de/conferences/ecer-2026-tampere (accessed on 16 March 2026).
[14] Eurostat (2025), Eurostat Database, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Adult_Education_Survey_(AES)_methodology.
[34] Eurostat (2022), EU Adult Education Survey 2022, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Adult_Education_Survey_(AES)_methodology (accessed on 5 May 2022).
[6] Eurostat (2019), Harmonised European Time Use Survey (HETUS) 2018 Guidelines, European Commission, https://ec.europa.eu/eurostat/documents/3859598/9710775/KS-GQ-19-003-EN-N.pdf/ee48c0bd-7287-411a-86b6-fb0f6d5068cc?t=1554468617000.
[8] Eurostat (2017), 2016 AES manual: Methodological guidelines for the Adult Education Survey, European Commission, Directorate F: Social Statistics, Unit F-3: Labour market and lifelong learning, https://ec.europa.eu/eurostat.
[7] Eurostat (2016), Classification of Learning Activities (CLA), Manual 2016 edition, https://ec.europa.eu/eurostat/documents/3859598/7659750/KS-GQ-15-011-EN-N.pdf.
[13] Groves, R. et al. (2004), Survey Methodology, New York: Wiley.
[35] Institute for Employment Research (2026), ASCOT: Computer Assisted Structured Coding Tool, University of Warwick, https://warwick.ac.uk/fac/soc/ier/data_group/cascot/.
[23] Kalenda, J. and E. Boeren (2025), “Revisiting the triadic classification of learning activities: rethinking their measurement”, International Journal of Lifelong Education, Vol. 44/3, pp. 257-272, https://doi.org/10.1080/02601370.2025.2489506.
[25] Karhapää, A., R. Hämäläinen and J. Pöysä-Tarhonen (2025), “Digital work practices that promote informal workplace learning: digital ethnography in a knowledge work context”, Studies in Continuing Education, Vol. 47/1, pp. 1-18, https://doi.org/10.1080/0158037X.2023.2274596.
[29] Karhapää, A., R. Hämäläinen and J. Pöysä-Tarhonen (2025), “Digital work practices that promote informal workplace learning: digital ethnography in a knowledge work context”, Studies in Continuing Education, Vol. 47/1, pp. 1-18, https://doi.org/10.1080/0158037X.2023.2274596.
[39] Koopmans, L. et al. (2025), “Informal learning and innovative employee behaviour”, TNO Public.
[37] Kunn, A. et al. (2024), “Learning and development in the Netherlands”, ROA Reports No. 005, https://cris.maastrichtuniversity.nl/en/publications/leren-en-ontwikkelen-in-nederland/.
[38] Kunn-Nelen, A. et al. (2025), “Exploring datasets for an LLO Radar”, ROA Technical Report, https://cris.maastrichtuniversity.nl/en/publications/verkenning-datasets-voor-een-llo-radar/.
[33] Littlejohn, A., E. Kennedy and D. Laurillard (2022), “Professional Learning Analytics: Understanding Complex Learning Processes Through Measurement, Collection, Analysis, and Reporting of MOOC Data”, https://doi.org/10.1007/978-3-031-08518-5_25.
[17] Mellander, E. (2017), “On the use of register data in educational science research”, Nordic Journal of Studies in Educational Policy, Vol. 3/1, pp. 106-118, https://doi.org/10.1080/20020317.2017.1313680.
[18] OECD (2025), Survey of Adult Skills (PIAAC), http://www.oecd.org/skills/piaac/ (accessed on 25 February 2019).
[10] OECD (2024), Survey of Adult Skills – Reader’s Companion: 2023, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/3639d1e2-en.
[9] OECD (2011), PIAAC Conceptual Framework of the Background, https://piaac.cz/wp-content/uploads/PIAAC2011_11MS_BQ_ConceptualFramework_1-Dec-2011.pdf.
[4] OECD (2021a), Computer-based Student Questionnaire for PISA 2022, OECD Publishing, https://www.oecd.org/content/dam/oecd/en/data/datasets/pisa/pisa-2022-datasets/questionnaires/COMPUTER-BASED%20STUDENT%20questionnaire%20PISA%202022.pdf.
[5] OECD (2021b), ICT Questionnaire for PISA 2022, OECD Publishing, https://www.oecd.org/content/dam/oecd/en/data/datasets/pisa/pisa-2022-datasets/questionnaires/ICT%20QUESTIONNAIRE%20PISA%202022.pdf.
[30] Oleaga, M. (2025), “Knowledge exchange in peripheral coworking spaces: A study of proximities using social network analysis”, Geoforum, Vol. 161, p. 104264, https://doi.org/10.1016/j.geoforum.2025.104264.
[20] ROA (2024), Lifelong Learning and Development, https://roa.nl/projects/lifelong-learning-and-development (accessed on 30 September 2025).
[15] Robinson, S. and K. Leonard (2024), Designing quality survey questions, SAGE Publications.
[36] SER (2024), Monitor Leercultuur 2023, https://publicaties.ser.nl/ser_leercultuurmonitor_2023/cover (accessed on 16 October 2025).
[19] SER (2024), Montior Leercultuur 2023, https://publicaties.ser.nl/ser_leercultuurmonitor_2023/cover (accessed on 16 October 2025).
[27] Smyth, E. (2022), The changing social worlds of 9-year-olds, Economic and Social Research Institute, Dublin, Ireland, https://doi.org/10.26504/rs151.
[24] Smyth, E. (2022), “The Changing Social Worlds of 9-Year-Olds”, ESRI Research Series, p. No 151, https://www.esri.ie/system/files/publications/RS151.pdf.
[40] The UN Network of Economic Statisticians (2024), Bureau Briefing Paper: Implementing the Expert Group on Wellbeing Measurement, UN, https://unstats.un.org/unsd/statcom/groups/EGWM/Meetings/egwm-FirstMeeting/Session4-1-Bureau-Briefing-Paper-Implementing-the-Expert-Group-on-Wellbeing-Measurement.pdf?utm_source=.
[31] Van Waes, S. and K. Hytönen (2022), “A Social Network Perspective on Workplace Learning and Professional Development”, https://doi.org/10.1007/978-3-030-89582-2_8.
[3] Verian (2024), European Training and Learning Survey: Technical and quality report, CEDEFOP, https://www.cedefop.europa.eu/files/etls_-_technical_and_quality_report_-_october_2024_0.pdf.
[11] Widany, S. et al. (2019), “The Quality of Data on Participation in Adult Education and Training. An Analysis of Varying Participation Rates and Patterns Under Consideration of Survey Design and Measurement Effects”, Frontiers in Sociology, Vol. 4, https://doi.org/10.3389/fsoc.2019.00071.