6. Accuracy and reliability |
Top |
|
|
6.1. Accuracy - overall |
[not requested for the LFS quality report] |
6.2. Sampling error |
Publication thresholds |
Annual estimates |
Annual estimates - wave approach |
(if different from full sample thresholds) |
Limit below which figures cannot be published |
Limit below which figures must be published with warning |
Limit below which figures cannot be published |
Limit below which figures must be published with warning |
500 |
Limit is 2 500. The explanation is; "sample size is too small for reliable estimates for figures less than two thousand persons in each cell". |
NA |
NA |
|
6.2.1. Sampling error - indicators |
Coefficient of variation (CV) Annual estimates Sampling error - indicators - Coefficient of variation (CV), Standard Error (SE) and Confidence Interval (CI) |
|
Number of employed persons |
Employment rate as a percentage of the population |
Number of part-time employed persons |
Number of unemployed persons |
Unemployment rate as a percentage of labour force |
Youth unemployment rate as a percentage of labour force |
Average actual hours of work per week(*) |
|
Age group: 20 - 64 |
Age group: 20 - 64 |
Age group: 20 - 64 |
Age group: 15 -74 |
Age group: 15 -74 |
Age group: 15 -24 |
Age group: 20 - 64 |
CV |
0.004953 |
0.31 |
0.01 |
0.01 |
1.03 |
1.49 |
0.00 |
SE |
123563.73 |
0.16 |
27841.00 |
45555.00 |
0.14 |
0.38 |
0.07 |
CI(**) |
(24706385-25190755) |
(50,744-51,369) |
(2708742-2817881) |
(3971674-4150251) |
(12,92-13,46) |
(24,53-26,02) |
(43,20-43,49) |
Description of the assumption underlying the denominator for the calculation of the CV for the employment rate |
In Turkstat for employment totals and rates, we have different cv values. We apply proc survey means sas module which is based on Taylor approach. We take into account the calculation formulas of standard deviation formula for the total estimates and the standard error formula for the mean estimates of the indicators required in the table. |
Reference on software used: |
Reference on method of estimation: |
SAS Enterprise Guide 5.1 |
TAYLOR SERIES |
Coefficient of variation (CV) Annual estimates at NUTS-2 Level |
NUTS-2 |
CV of regional (NUTS-2) annual aggregates (in %) |
Regional Code |
Region |
Number of employed persons |
Employment rate as a percentage of the population |
Number of part-time employed persons |
Number of unemployed persons |
Unemployment rate as a percentage of labour force |
Youth unemployment rate as a percentage of labour force |
Average actual hours of work per week(*) |
|
|
Age group: 20 - 64 |
Age group: 20 - 64 |
Age group: 20 - 64 |
Age group: 15 -74 |
Age group: 15 -74 |
Age group: 15 -24 |
Age group: 20 - 64 |
TR10 |
İstanbul |
1.18 |
0.83 |
2.61 |
2.94 |
2.68 |
4.34 |
0.44 |
TR21 |
Tekirdağ, Edirne, Kırklareli |
2.66 |
1.26 |
5.09 |
5.37 |
4.84 |
7.37 |
0.69 |
TR22 |
Balıkesir, Çanakkale |
2.55 |
1.56 |
5.01 |
6.19 |
6.24 |
9.47 |
0.81 |
TR31 |
İzmir |
2.31 |
1.39 |
3.45 |
3.72 |
3.51 |
4.65 |
0.61 |
TR32 |
Aydın, Denizli, Muğla |
1.97 |
1.23 |
3.77 |
5.40 |
5.10 |
8.89 |
0.78 |
TR33 |
Manisa, Afyon, Kütahya, Uşak |
2.62 |
1.29 |
4.39 |
5.43 |
5.07 |
7.10 |
0.62 |
TR41 |
Bursa, Eskişehir, Bilecik |
1.89 |
1.23 |
3.67 |
4.71 |
4.44 |
6.67 |
0.61 |
TR42 |
Kocaeli, Sakarya, Düzce, Bolu, Yalova |
1.77 |
0.98 |
3.77 |
4.43 |
3.69 |
5.06 |
0.69 |
TR51 |
Ankara |
1.65 |
1.07 |
4.15 |
3.87 |
3.51 |
4.51 |
0.63 |
TR52 |
Konya, Karaman |
2.66 |
1.56 |
5.28 |
4.55 |
4.96 |
6.69 |
0.71 |
TR61 |
Antalya, Isparta, Burdur |
2.09 |
1.18 |
3.99 |
4.86 |
4.59 |
6.65 |
0.87 |
TR62 |
Adana, Mersin |
2.48 |
1.44 |
4.16 |
5.19 |
4.84 |
6.34 |
0.80 |
TR63 |
Hatay, Kahramanmaraş, Osmaniye |
2.64 |
1.72 |
4.50 |
4.60 |
4.59 |
5.82 |
0.77 |
TR71 |
Kırıkkale, Aksaray, Niğde, Nevşehir, Kırşehir |
2.82 |
1.64 |
4.58 |
4.32 |
4.40 |
6.48 |
0.76 |
TR72 |
Kayseri, Sivas, Yozgat |
2.91 |
1.54 |
6.98 |
5.64 |
4.87 |
6.43 |
1.04 |
TR81 |
Zonguldak, Karabük, Bartın |
2.73 |
1.90 |
5.95 |
7.29 |
6.81 |
8.79 |
0.82 |
TR82 |
Kastamonu, Çankırı, Sinop |
4.88 |
2.60 |
6.59 |
10.50 |
11.02 |
10.43 |
1.03 |
TR83 |
Samsun, Tokat, Çorum, Amasya |
2.94 |
1.67 |
5.85 |
5.35 |
5.41 |
7.87 |
1.00 |
TR90 |
Trabzon, Ordu, Giresun, Rize, Artvin, Gümüşhane |
2.50 |
1.19 |
4.80 |
5.31 |
4.99 |
7.31 |
0.68 |
TRA1 |
Erzurum, Erzincan, Bayburt |
3.65 |
2.19 |
6.41 |
7.59 |
7.50 |
9.26 |
0.94 |
TRA2 |
Ağrı, Kars, Iğdır, Ardahan |
2.57 |
2.33 |
4.65 |
7.04 |
6.51 |
6.09 |
0.97 |
TRB1 |
Malatya, Elazığ, Bingöl, Tunceli |
3.25 |
2.18 |
6.27 |
5.95 |
6.28 |
9.64 |
0.77 |
TRB2 |
Van, Muş, Bitlis, Hakkari |
4.62 |
2.62 |
8.51 |
5.26 |
4.67 |
5.72 |
1.33 |
TRC1 |
Gaziantep, Adıyaman, Kilis |
2.45 |
1.69 |
6.35 |
6.47 |
5.92 |
9.02 |
0.83 |
TRC2 |
Şanlıurfa, Diyarbakır |
3.31 |
2.24 |
8.53 |
5.42 |
4.64 |
9.27 |
1.25 |
TRC3 |
Mardin, Batman, Şırnak, Siirt |
4.34 |
3.23 |
7.61 |
6.83 |
5.02 |
4.34 |
1.37 |
(*) The coefficient of variation for actual hours worked should be calculated for the sum of actual hours worked in 1st and 2nd jobs, and restricted to those who actually worked 1 hour or more in the reference week. (**) The value is based on a CI of 95%. For the rates the CI should be given with 2 decimals. |
6.3. Non-sampling error |
[not requested for the LFS quality report] |
6.3.1. Coverage error |
Frame quality (under-coverage, over-coverage and misclassifications(b)) |
Under-coverage rate (%) |
Over-coverage rate (%) |
Misclassification rate (%) |
Comments: specification and impact on estimates(a) |
|
Undercoverage |
Overcoverage |
Misclassification(b) |
Reference on frame errors |
UNA |
UNA |
UNA |
UNA |
UNA |
UNA |
UNA |
(a) Mention specifically which regions / population groups are not suitably represented in the sample. (b) Misclassification refers to statistical units having an erroneous classification where both the wrong and the correct one are within the target population. |
6.3.1.1. Over-coverage - rate |
[Over-coverage rate, please see concept 6.3.1 Coverage error in the LFS quality report] |
6.3.1.2. Common units - proportion |
[not requested for the LFS quality report] |
6.3.2. Measurement error |
Errors due to the medium (questionnaire) |
Was the questionnaire updated for the 2020 LFS operation? (Y/N) |
Synthetic description of the update |
Was the questionnaire tested? (Y/N) |
If the questionnaire has been tested, which kind of tests has been applied (pilot, cognitive, internal check)? |
Y |
In order to estimate the impact of the new regulations and IESS on the indicators and use it in comparable back-calculation of the series, questions were added to the existing questionnaire for measuring the definition effect in 2020. |
Y |
Pilot, cognitive and internal tests were made. |
Main methods of reducing measurement errors |
Error source |
|
Respondent |
Letter introducing the survey (Y/N) |
Phone call for booking or introducing the survey (Y/N) |
Y |
N |
Interviewer |
Periodical training (at least 1 time per year) (Y/N) |
Feedbacks from interviewer (reports, debriefings, etc.) (Y/N) |
Y |
Y |
Fieldwork |
Monitoring directly by contacting the respondents after the fieldwork (Y/N) |
Monitoring directly by listening the interviews (Y/N) |
Monitoring remotely through performance indicators (Y/N) |
Y |
Y |
Y |
Questionnaire |
Questionnaire in several languages (Y/N) |
On-line checks (for computer assisted interviews (Y/N) |
Y |
Y |
Other / Comments |
Quality control study is regularly realized almost every month of year for around 1500 households which were intervieved in the previous week. In this study; households which are selected randomly from each region and interviewer were called and some critical questions were again asked to the selected household member. This is a very good tool for minimazing measuremet errors in the later months. |
|
6.3.3. Non response error |
[not requested for the LFS quality report] |
6.3.3.1. Unit non-response - rate |
IN THIS SECTION INFORMATION REFERS TO THE FINAL SAMPLING UNITS *
Methods used for adjustments for statistical unit non-response |
Adjustment via weights (Y/N) |
Variables used for non-response adjustment |
Description of method |
Y |
Household nonresponse is adjusted on the basis of blocks |
The adjustment is made in each block by the formula 10/r where r is the number of respondent household |
Substitution of non-responding units (Y/N) |
Substitution rate |
Criteria for substitution |
N |
NA |
NA |
Other methods (Y/N) |
Description of method |
N |
NA |
Non-response rates by survey mode. Annual average (% of the theoretical yearly sample by survey mode) |
Survey |
CAPI |
CATI |
PAPI |
CAWI |
POSTAL |
5.02 |
6.55 |
NA |
NA |
NA |
Divisions of non-response into categories. Quarterly data and annual average |
Quarter |
Non-response rate |
Total (%) |
of which: |
Refusals (%) |
Non-contacts (including people who migrated (or moved) internally or abroad) (%) |
of which people who migrated (or moved) internally or abroad (%) |
1 |
5.31 |
5.33 |
79.29 |
5.36 |
2 |
4.66 |
19.22 |
66.99 |
4.36 |
3 |
5.77 |
11.84 |
75.68 |
4.23 |
4 |
6.07 |
9.55 |
76.58 |
4.30 |
Annual |
5.46 |
11.24 |
74.92 |
4.56 |
Units who refused to participate in the survey (Please indicate the number of the units concerned in the cells where the wave is mentioned) |
Subsample |
Quarter1_2020 |
Quarter2_2020 |
Quarter3_2020 |
Quarter4_2020 |
Subsample_Q4_2018 |
35 |
|
|
|
Subsample_Q1_2019 |
14 |
54 |
|
|
Subsample_Q2_2019 |
|
40 |
55 |
|
Subsample_Q3_2019 |
|
|
30 |
36 |
Subsample_Q4_2019 |
31 |
|
|
33 |
Subsample_Q1_2020 |
60 |
140 |
|
|
Subsample_Q2_2020 |
|
223 |
189 |
|
Subsample_Q3_2020 |
|
|
76 |
192 |
Subsample_Q4_2020 |
|
|
|
35 |
Total in absolute numbers |
140 |
457 |
350 |
296 |
Total in % of theoretical quarterly sample |
0.24 |
0.78 |
0.60 |
0.51 |
Units who were not contacted (including people who migrated (or moved) internally or abroad) (Please indicate the number of units only in the cells where the wave is mentioned) |
Subsample |
Quarter1_2020 |
Quarter2_2020 |
Quarter3_2020 |
Quarter4_2020 |
Subsample_Q4_2018 |
316 |
|
|
|
Subsample_Q1_2019 |
247 |
275 |
|
|
Subsample_Q2_2019 |
|
220 |
394 |
|
Subsample_Q3_2019 |
|
|
243 |
467 |
Subsample_Q4_2019 |
431 |
|
|
275 |
Subsample_Q1_2020 |
1089 |
641 |
|
|
Subsample_Q2_2020 |
|
457 |
1120 |
|
Subsample_Q3_2020 |
|
|
480 |
1184 |
Subsample_Q4_2020 |
|
|
|
447 |
Total in absolute numbers |
2083 |
1593 |
2237 |
2373 |
Total in % of theoretical quarterly sample |
3.56 |
2.72 |
3.82 |
4.05 |
of which people who migrated (or moved) internally or abroad) (Please indicate the number of units only in the cells where the wave is mentioned) |
Subsample |
Quarter1_2020 |
Quarter2_2020 |
Quarter3_2020 |
Quarter4_2020 |
Subsample_Q4_2018 |
285 |
|
|
|
Subsample_Q1_2019 |
911 |
324 |
|
|
Subsample_Q2_2019 |
|
870 |
261 |
|
Subsample_Q3_2019 |
|
|
876 |
441 |
Subsample_Q4_2019 |
506 |
|
|
1121 |
Subsample_Q1_2020 |
7415 |
688 |
|
|
Subsample_Q2_2020 |
|
5562 |
596 |
|
Subsample_Q3_2020 |
|
|
5504 |
916 |
Subsample_Q4_2020 |
|
|
|
4864 |
Total in absolute numbers |
9117 |
7444 |
7237 |
7342 |
Total in % of theoretical quarterly sample |
15.57 |
12.71 |
12.36 |
12.54 |
Non-response rates. Annual averages (% of the theoretical yearly sample) |
NUTS-2 region (code + name) |
Non response rate (%) |
TR10-İstanbul |
3.76 |
TR21-Tekirdağ, Edirne, Kırklareli |
5.63 |
TR22-Balıkesir, Çanakkale |
5.54 |
TR31-İzmir |
7.04 |
TR32-Aydın, Denizli, Muğla |
4.66 |
TR33-Manisa, Afyon, Kütahya, Uşak |
4.73 |
TR41-Bursa, Eskişehir, Bilecik |
4.08 |
TR42-Kocaeli, Sakarya, Düzce, Bolu, Yalova |
5.86 |
TR51-Ankara |
2.81 |
TR52-Konya, Karaman |
5.34 |
TR61-Antalya, Isparta, Burdur |
5.23 |
TR62-Adana, Mersin |
5.19 |
TR63-Hatay, Kahramanmaraş, Osmaniye |
5.38 |
TR71-Kırıkkale, Aksaray, Niğde, Nevşehir |
5.96 |
TR72-Kayseri, Sivas, Yozgat |
3.09 |
TR81-Zonguldak, Karabük, Bartın |
5.16 |
TR82-Kastamonu, Çankırı, Sinop |
5.53 |
TR83-Samsun, Tokat, Çorum, Amasya |
4.58 |
TR90-Trabzon, Ordu, Giresun, Rize, Artvin |
5.88 |
TRA1-Erzurum, Erzincan, Bayburt |
6.36 |
TRA2-Ağrı, Kars, Iğdır, Ardahan |
11.38 |
TRB1-Malatya, Elazığ, Bingöl, Tunceli |
5.15 |
TRB2-Van, Muş, Bitlis, Hakkari |
5.50 |
TRC1-Gaziantep, Adıyaman, Kilis |
6.87 |
TRC2-Şanlıurfa, Diyarbakır |
11.97 |
TRC3-Mardin, Batman, Şırnak, Siirt |
5.56 |
* If the final sampling unit is the household it must be considered as responding unit even in case of some household members (not all) do not answer the interview |
6.3.3.2. Item non-response - rate |
Item non-response (*) - Quarterly data (Compared to the variables defined by the Commission Regulation (EC) No 377/2008) |
Variable status |
Column |
Identifier |
Quarter 1 |
Quarter 2 |
Quarter 3 |
Quarter 4 |
Short comments on reasons for non-available statistics and prospects for future solutions |
Compulsory / optional |
compulsory |
Col_017/18 |
NATIONAL |
100.0 |
100.0 |
100.0 |
100.0 |
According to the 2010 Address Based Population Registration System; 99,7 % of population has Turkish Nationality. So, it is not easy to cover non-nationals with a sample survey. |
compulsory |
Col_039/40 |
COUNTRYW |
C |
C |
C |
C |
There are very few people who are working abroad and at the same time considered as household member since Turkey is a very broad country. This may only occur in border cities, but not common. So, this variable is not asked. |
compulsory |
Col_067/68 |
HWOVERPU |
100.0 |
100.0 |
100.0 |
100.0 |
Only total overtime is asked in the questionnaire (paid+unpaid). Since it is not possible to distinguish paid and unpaid overtime. Total overtime is given in HWOVERP and this variable is coded as blank. |
compulsory |
Col_073/74 |
HWWISH |
100.0 |
100.0 |
100.0 |
100.0 |
This variable was dropped out from the questionnaire in 2009 since it was observed that, results were not reliable. Respondents replied this question as they undestand (some give the hours that would like to work in total while others only give the additional hours). |
compulsory |
Col_104 - Employed |
METHODB |
. |
C |
C |
C |
For employed people all the methods are not asked in same detail, some of them are grouped looking at the frequency (for example, Public Employment Office and Private Employment Offices are combined in one code). |
compulsory |
Col_108 - Employed |
METHODF |
. |
C |
C |
C |
For employed people all the methods are not asked in same detail, some of them are grouped looking at the frequency (for example, Public Employment Office and Private Employment Offices are combined in one code). |
compulsory |
Col_111 - Employed |
METHODI |
. |
C |
C |
C |
For employed people all the methods are not asked in same detail, some of them are grouped looking at the frequency (for example, Public Employment Office and Private Employment Offices are combined in one code). |
compulsory |
Col_113 - Employed |
METHODK |
. |
C |
C |
C |
For employed people all the methods are not asked in same detail, some of them are grouped looking at the frequency (for example, Public Employment Office and Private Employment Offices are combined in one code). |
compulsory |
Col_114 - Employed |
METHODL |
. |
C |
C |
C |
For employed people all the methods are not asked in same detail, some of them are grouped looking at the frequency (for example, Public Employment Office and Private Employment Offices are combined in one code). |
compulsory |
Col_115 - Not employed |
METHODM |
C |
. |
. |
. |
|
compulsory |
Col_168 |
DEGURBA |
100.0 |
100.0 |
100.0 |
100.0 |
We are planning to transmit this variable to Eurostat from 2021 on. |
Item non-response - Annual data (Compared to the variables defined by the Commission Regulation (EC) No 377/2008) |
Variable status |
Column |
Identifier |
This reference year |
Short comments on reasons for non-available statistics and prospects for future solutions |
compulsory |
Col_055 |
TEMPAGCY |
100 |
This variable is not asked since tempopary working agencies are not common in Turkey for the moment. |
compulsory |
Col_121 |
REGISTER |
100 |
This question is not asked since the coverage of unemployment benefits is very limited in Turkey. (around 10% of registered unemployed are receiving unemployment benefit at the current situation). |
optional |
Col_132 |
COURPURP |
100 |
Questions about attending any courses, seminars, conferences or receive private lessons or instructions outside the regular education system haven't asked since 2014. |
optional |
Col_133/135 |
COURFILD |
100 |
Questions about attending any courses, seminars, conferences or receive private lessons or instructions outside the regular education system haven't asked since 2014. |
optional |
Col_136 |
COURWORH |
100 |
Questions about attending any courses, seminars, conferences or receive private lessons or instructions outside the regular education system haven't asked since 2014. |
(*) "C" means all the records have the same value different from missing. |
6.3.4. Processing error |
Editing of statistical item non-response |
Do you apply some data editing procedure to detect and correct errors? (Y/N) |
Overall editing rate (Observations with at least one item changed / Total Observations ) |
N |
NA |
|
6.3.4.1. Imputation - rate |
Imputation of statistical item non-response |
Are all or part of the variables with item non response imputed? (Y/N) |
Overall imputation rate (Observations with at least one item imputed / Total Observations ) |
N |
NA |
Main variables |
Imputation rate |
Describe method used, mentioning which auxiliary information or stratification is used |
NA |
NA |
NA |
|
6.3.5. Model assumption error |
[not requested for the LFS quality report] |
6.4. Seasonal adjustment |
Do you apply any seasonal adjustment to the LFS Series? (Y/N) |
If Yes, is your adopted methodology compliant with the ESS guidelines on seasonal adjustment? (ref. http://ec.europa.eu/eurostat/web/research-methodology/seasonal-adjustment) (Y/N) |
If Yes, are you compliant with the Eurostat/ECB recommendation on Jdemetra+ as software for conducting seasonal adjustment of official statistics. (ref. http://ec.europa.eu/eurostat/web/ess/-/jdemetra-officially-recommended-as-software-for-the-seasonal-adjustment-of-official-statistics) (Y/N) |
If Not, please provide a description of the used methods and tools |
Y |
Y |
N |
Seasonal adjustment of Labour Force Statistics carries out by using TRAMO-SEATS methodology based on ARIMA (Autoregressive Integrated Moving Average) model estimation developed by the Banco de Espana and also suggested by Eurostat. The software that is used for the application of this method is TRAMO-SEATS for Windows (TSW). Seasonally adjusted figures of labour force statistics have been produced by indirect approach. Namely, labour force, employed persons according to economic activities and unemployed persons are firstly seasonally adjusted and then aggregated to derive seasonally adjusted employment and unemployment rates. |
|
6.5. Data revision - policy |
|
6.6. Data revision - practice |
[not requested for the LFS quality report] |
6.6.1. Data revision - average size |
[not requested for the LFS quality report] |