Final database

• Pages: 32-34

Final datab ase

European Inst itute for Gender Equalit y

32

8. Final database

(18) Reques ts should b e sent to eige.sec@eige.europa.eu

8.1. Micro and macro data

The analysis of results may be carried out

according to gender (female, male) and age

group (18-39, 40-64, and ≥ 65) at European level

and at cluster level, with the exception of some

cases in Q9 and Q10.

Table 1 provides information on the sample

sizes, both planned and nal. Due to the lter

question (Q8), some variables (Q9, Q10, Q11)

have insucient data to perform the analysis by

cluster or Member State within an acceptable

error level.

Annex 4 provides the sampling errors of a pro-

portion of each question at Member State, clus-

ter and EU-28 level, highlighting in green the

cases with a sampling error that could be con-

sidered acceptable (≤ 5 %).

In some exceptional cases where the sample size

is adequate for some items, depending on gen-

der, suciently precise estimates could be cal-

culated at Member State level. A rule of thumb

could be to accept a sample size where it has at

least 50 respondents in each category of interest.

The fact that this database only provides infor-

mation for people with direct experience of the

services evaluated is not a drawback. Rather,

these data are highly valuable, as evaluations

of the services are based on personal experi-

ence rather than on others experience, word of

mouth or intuitive inferences. Users with direct

experience of the services analysed are likely

to have clear perceptions and thus provide an

accurate evaluation of the services. Drawing

conclusions from the overall population is not

possible, but drawing conclusions from service

users certainly is.

The database at cluster and EU level is availa-

ble on request in SPSS format (.sav) (

18

). The le

includes all of the questions in the questionnaire,

accompanied by additional contextual informa-

tion (see Section 9 for description). The aggrega-

tion of the scale and ordinal variables was done

using the arithmetic mean, while the nominal

variables (Q8, Q14 and Q16) were aggregated

through the median.

8.2. Contextual information

Perceptions of context are useful where indi-

viduals elaborate on the subjective perceptions

that an ad hoc survey aims to measure. To facil-

itate testing the impact of the context, the data-

base includes a set of macro indicators that aim

to reconstruct the economic, health and polit-

ical environments of each cluster, with special

reference to gender inequality.

That system of context indicators (SCI) was

compiled through dierent data sources and is

structured according to the following indicator

areas (see Annex 5):

• income and income distribution

• labour market

• gender

• transport

• social expenditure

• crime and corruption

• environment

• culture

• violence against women.

Income and income distribution

Income and income distribution contain GDP at

market prices and per capita GDP, which allows

for a measure of the income of each Member

State. However, specic indicators have been

included on income distribution (Gini Index) and

poverty (money poverty, material deprivation