|
?:comment
|
-
Note:
/ SI: The database includes a post-stratification weight, which equalizes the structural shares of the obtained sample with the population shares derived from statistical sources (sample frame, source: National statistical office).
The weight is based on a combination of variables: Gender x Age (2x4) and Statistical region x Type of settlement (12x6).
/ ZA: Separate weighting factors for the design weight will be delivered as soon as possible (see errata).
(en)
-
PSWEIGHT: Post-stratification weight
This post-stratification weight serves as compensation for possible biases introduced by unit non-response or non-random sampling error that may occur during data collection. Post-stratification weights are usually calculated by comparing the relative frequency distribution of certain (usually demographic, such as age, gender, education) characteristics as found in the sample to a known reference distribution of the same characteristics, reported e.g. in recent official statistics.
It is not mandatory that ISSP members provide such weights, but if such weights are present, they are to be offered in this variable. ISSP users may want to enable this weight if they want sample distributions that match population reference distributions on demographic core characteristics. Whether this kind of weighting is desirable or not is debated in the methodological literature and is therefore left to the informed choice of the users.
COMMENT ON PSWEIGHT: PSWEIGHT should be computed such that the sum of the individual case weights equals the case count of the unweighted net sample (=number of cases in the data set). If no poststratification weight was provided, the variable PSWEIGHT must be set to '1'.
Detailed reporting about construction of the post-stratification weight is mandatory and must be provided with the national Technical Report.
...............................................................................................................................................................
AT: Yes, the post-stratification weight adjusts for: age, gender, education and age x education.
AU: Yes, post-stratification and non-response; the post-stratification weight adjusts for: age, sex and education.
CH: No.
CN: Yes, post-stratification and non-response weight; post-stratification weights adjust for: gender, age, education, urbanization.
CZ: Yes, the post-stratification weight adjusts for: SIZE6AM x REG (size of resident's site x region), SEX, AGE (age category), CZ_ISCD (B2, education category), WORK (B04, working status)
DE: No.
DK: Yes, based on GENDER, AGE in four groups (18-34/ 35-49/ 50-64 and 65+), ETHNICITY in two groups (Danish origin/ Immigrants and remnants), FAMILY TYPE in four groups (Single without children/ Single with children/ Couple without children/ Couple with children), EDUCATION in three groups (Primary school and missing/ General upper secoundary school, vocational education, combined general and vocational upper secoundary school, short-cycle tertiray education/ University bachelor degree, medium and long tertiary education, researcher, officer and masters course), SOCIOECONOMIC STATUS in three groups (Employed/ Retired/ Other).
FI: Yes, post-stratification weight provided based on gender, age, degree of urbanization, and region.
FR: Yes, post stratification weighting computed on sex, age (4 groups : 18 to 29 years old, 30 to 39 years old, 40 to 54 years old, 55 years old or more), occupation (6 groups: Farmers; Tradesmen, Shopkeepers and Business Owners; Managers and Secondary/ University Teachers; Intermediate Professions, White Collar Workers; Blue Collar Workers; Unemployed) and education level (4 levels).
HR: Post-stratification weight provided based on region, settlement size, gender and age.
HU: Yes, post-stratification weights used based on sex, age, education and type of settlement.
IL: Yes, post-stratification weight used based on the variables c_sample(Jews/ Arabs)*SEX*EDUCYRS.
IN: Yes, multiple weighting done on the following variables: Age, Gender, Social status, Income, Education.
IS: No.
IT: No.
JP: No.
MX: No.
NL: No.
NO: No.
NZ: Yes, post-stratification weights were based on logistic regression models of Responded (yes/no) by age group, Māori descent, occupation, region, deprivation quintile, and urbanicity. This was not to correct directly for the disproportionate sampling of the stratification, but it did this in the process of adjusting the sample back closer to the distributions of these variables in the initial sample from the New Zealand electoral rolls. The stratification, especially the oversampling of minority ethnic groups, was designed to account for known lower response rates among these groups - the primary aim was to achieve sufficient responses from these groups to allow analysis by ethnicity to be conducted.
PH: Yes, a post-stratification weight was used which is based on people aged 18+ (voting age) and area (NCR, Balance Luzon: North/ Central and South Luzon, Visayas, Mindanao).
PL: Yes, based on voivodship (16), age group by sex (10), type and size of the city (4).
RU: Yes, post-stratification weights used based on sex, age, education.
SI: Yes.
SK: Yes, based on the variables/ information on gender, age, education, region and size of municipality.
SR: Yes, post-stratification weights used (based on stratum, gender, age, education, main activity, position in household).
TH: No.
TW: Yes, based on sex, age, urbanization and education-degree groups.
US: Yes, post-stratification weight provided based on region, marital status, hispanic ethnicity, education, US born, gender, race and age.
ZA: Yes, but currently there is no separate weighting factor for the post-stratification, only a combined weight (see WEIGHT_COM and errata).
(en)
|
![[This page as RDF]](https://data.gesis.org/gesiskg/static/rdf-icon.gif){kind=icon}