Awareness of the European Qualifications Framework

Abstract

The European Qualifications Framework (EQF) has been introduced as a common reference framework for communicating the equivalence of the different countries’ qualifications across Europe. It defines eight levels of education, from primary (at Level 1) to doctoral (at Level 8). The Europass CV online editor gives users the option to include their EQF level on their CV either by selecting it via a drop-down menu or by writing free text that describes it or an equivalent national qualification.

Data processing

EQF level on Europass CV data

eqfYear <- fread(getSourcePath("input/eqf_analysis/cv_generation_year_eqf_level_count.tsv"), quote = "", na.strings = "NULL")
eqfYear <- eqfYear[is.na(V4) & !is.na(count)][!grepl("[^0-9]", count)][, V4 := NULL]
eqfYear[, count := as.numeric(count)]
eqfYear %>% setnames(c("cv_year", "eqf_selected", "count"))

eqfCountry <- fread(getSourcePath("input/eqf_analysis/country_eqf_level_count_fix.tsv"), quote = "", na.strings = "NULL")
eqfCountry <- eqfCountry[is.na(V4) & !is.na(count)][!grepl("[^0-9]", count)][, V4 := NULL]
eqfCountry[, count := as.numeric(count)]
eqfCountry %>% setnames(c("country", "eqf_selected", "count"))

eqfAgeGroup <- fread(getSourcePath("input/eqf_analysis/age_group_eqf_level_count.tsv"), quote = "", na.strings = "NULL")
eqfAgeGroup <- eqfAgeGroup[is.na(V4) & !is.na(count)][!grepl("[^0-9]", count)][, V4 := NULL]
eqfAgeGroup[, count := as.numeric(count)]
eqfAgeGroup %>% setnames(c("age_group", "eqf_selected", "count"))

eqfLocale <- fread(getSourcePath("input/eqf_analysis/doc_language_eqf_level_count.tsv"), quote = "", na.strings = "NULL")
eqfLocale <- eqfLocale[is.na(V4) & !is.na(count)][!grepl("[^0-9]", count)][, V4 := NULL]
eqfLocale[, count := as.numeric(count)]
eqfLocale %>% setnames(c("locale", "eqf_selected", "count"))

eqfInstitution <- fread(getSourcePath("input/eqf_analysis/organisation_country_eqf_level_count.tsv"), quote = "", na.strings = "NULL")
eqfInstitution <- eqfInstitution[is.na(V4) & !is.na(count)][!grepl("[^0-9]", count)][, V4 := NULL]
eqfInstitution[, count := as.numeric(count)]
eqfInstitution %>% setnames(c("institution_country", "eqf_selected", "count"))

total_count <- eqfYear[, count] %>% sum(na.rm = TRUE) %>% epas_format
  • The EQF level field is optional when filling a CV in the Europass online editor. Users who choose to fill it can either select their EQF level via a drop-down menu (options include numbers 1 through 8) or freely type any text on the same field.
  • Data relating the EQF level field usage with the year of CV creation, language on to which the CV was written, countries of residence and institution, and age group has been exported from the database.
  • Each observation is a single entry in the Education and Training section of the CV. The total number of observations acquired is 272.2M.
  • This report is based on data extracted directly from the database, and as such the deduplication step hasn’t been applied. This means that the absolute numbers reported represent aggregations on raw data as they exist in the database, and not necessarily the true number of qualifications provided by users.
  • A more in-depth analysis was subsequently performed using the pipeline developed for the Europass CV Insights Report. It is detailed on the respective section below.
  • Note that the classification algorithm used was developed in the context of the Europass CV collection between June and September 2019. Due to the relatively small amount of data, an unsupervised method was used. However, in the Europass Database that was integrated in the last days of this project, we observe that it might be plausible to develop a classifier with supervised learning. In particular, we can use the free-text provided by the users along with the EQF level code selected from the drop-down menu (2.5M examples).

Dimensions available to the analysis

In the context of this analysis, it has been elected to assess the EQF level field usage with respect to five key features. Usage statistics are derived by looking at one dimension at a time.

  • Qualifications between 2016 and 2020 are provided. More specifically, the period between August 2016 and June 2020 is covered.
  • A total of 31 document languages exist in the database. A user may select one language to write their CV.
  • The ten most used document languages are in order: English, Italian, Portuguese, Spanish, German, Romanian, French, Hungarian, Croatian and Greek.
  • There is a total of over 150 countries in the database, of which 30 have been found to provide at least 1,000,000 observations.
  • The ten countries with most observations are in order: Italy, Portugal, Spain, Romania, Germany, Hungary, Croatia, Greece, Slovenia and France.
eqfInstitution[, usage := ifelse(eqf_selected %in% as.character(1:8), "Drop-down", "Free-text")]
eqfInstitution[is.na(eqf_selected), usage := "None"]
eqfInstitution[is.na(institution_country), institution_country := "Not filled"]
eqfInstitution[!(institution_country %in% c(ISO_3166_1$Alpha_2, "EL", "UK", "EN", "Not filled")), institution_country  := "Free text"]
eqfInstitution[(institution_country %in% c(ISO_3166_1$Alpha_2, "EL", "UK", "EN")), institution_country := country_name(institution_country)]

eqfInstitution_use <- eqfInstitution[, .(institution_country, eqf_selected, usage, count)]
eqfInstitution_use[is.na(eqf_selected), usage := "None"]
eqfInstitution_use <- eqfInstitution_use[, .(count = sum(count, na.rm = TRUE)), by = c("institution_country", "usage")]
eqfInstitution_use[, usage := factor(usage, levels = c("Drop-down", "Free-text", "None"))]

eqfInstitution_level <- eqfInstitution[, .(count = sum(count, na.rm = TRUE)), by = c("institution_country", "eqf_selected", "usage")]
eqfInstitution_level[usage == "Free-text", eqf_selected := "Free text"]
eqfInstitution_level[usage == "None", eqf_selected := "Not filled"]
eqfInstitution_level <- eqfInstitution_level[, .(count = sum(count, na.rm = TRUE)), by = c("institution_country", "eqf_selected")]

top_institution_countries <- eqfInstitution[!(institution_country %in% c("Free text", "Not filled")), .(count = sum(count, na.rm = TRUE)), by = "institution_country"][order(-count), institution_country]
top_10_institution_countries <- top_institution_countries[1:9] %>% paste(collapse = ", ") %>% paste("and", top_institution_countries[10])
  • Likewise, the ten most reported institution countries are Italy, Portugal, Spain, Romania, Germany, Croatia, Hungary, France, Greece and Slovenia.
  • Ages have been split into five groups.
Usage percantage of the EQF level drop-down menu by age group.
Age group Total observations Usage percentage
Up to 20 16.7M 14%
21-25 42.1M 13%
26-30 34.8M 13%
31-35 23.7M 9%
36+ 34.4M 8%
Unknown 120.6M 11%

Resulting data sets

Usage statistics of the EQF level field by each respective dimension have been aggregated and compiled into simple downloadable datasets of .CSV format. They and can be accessed using the buttons bellow.

Year of CV creation Language used to complete CV

Country of residence Institution country Age group

Analysis results

Explore Data

  • Around 16% of all observations have been found to make use of the EQF level field. The majority of those are a numeric selection from the drop-down menu.
  • Out of all drop-down selections 41% are for a level in primary and secondary education.
  • Note that this EQF level distribution does not necessarily represent the EQF level of all respondents, as it is a distribution of reported levels for all qualifications. For example, it is possible for a respondent with EQF level 8 to include both of his or her PhD (Level 8) and primary school education (Level 1).

  • Percentage of inclusion of the EQF level on Europass CVs exhibits no major differences for CV created between 2016 and 2020.
  • For users that made use of the drop-down menu to select the EQF level of each of their qualifications, the distribution is almost the same year-on-year.

  • Among the 10 languages most used, CVs written in Portuguese and English display the highest use of the EQF level field.
  • The lowest use is for CVs written in Hungarian and Romanian.
  • For instances where the drop-down menu was used, the distribution differs based on the language. For example, CVs written in English include higher EQF levels than average, while CVs written in Portuguese include lower levels.

  • Among the 20 countries with most observations, CVs reporting Malta and Portugal as country of residence made most use of the EQF level menu.
  • Respondents from Hungary, Austria, and Romania made least use of it, generally leaving it unfilled.
  • Generally, users from most countries more commonly make use of the drop-down menu instead of including a free-text selection, but certain cases like Bulgaria, Romania and India free-text entries are used almost equally as often.
  • There is a high degree of variation in the distribution of the EQF level drop-down selection based on country of residence. For example, users from Belgium mostly report university-level education, while the opposite is true for users from Portugal and Malta.

  • The ten most highly reported institution countries display similar patterns as the respective countries of residence.
  • Distribution of EQF level based on usage of the drop-down menu is highest for education received in Greek and German institutions.

  • Usage of the EQF level menu is very similar between the different age groups, displaying slight increase for lower ages.
  • High level education peaks for users at around 30 years old, with younger and older users generally reporting lower EQF levels.

In-depth data exploration

This report represents the first phase of the analysis done on the awareness of the European Qualifications Framework by users of the Europass CV online editor. A more comprehensive analysis was subsequently performed using the analysis pipeline developed for the Europass Survey. It allows to further drill-down on data, inspect multiple dimensions at once, as well as get insight on the free text included by users in the EQF level label.

Text Analysis

Popular keywords on EQF level free-text
Popular phrases on EQF level free-text

  • Data deduplication as well as further cleansing has been performed, and each free-text qualification entry has been matched to an EQF level.
  • The methodology used to retrieve the EQF level is identical to the one developed for the Survey Report. Note, however, that due to the high volume of data a smaller source corpus is utilized in this case.
  • Whereas the present document reports statistics for individual qualification entries, data both for users of Europass, and individual qualifications can be accessed in the more in-depth analysis.
  • Results can be accessed via the Exploratory Data Tool. The respective datasets are placed in the EQF Analysis section of the Codebook index.

Codebook index

References

  • European qualifications framework (EQF). Project info
  • Cedefop. 2020. “Europass CV Insights Report: July - September 2019.” Report
  • Cedefop. 2020. Europass Data Science: Exploratory Data Tool

2016 - 2020