Description of sampling methodology (details of how sample designed)



Download 67.86 Kb.
Date21.05.2016
Size67.86 Kb.
2-5 page Word Document containing:
Description of sampling methodology (details of how sample designed)
For data collection we used the ’centres of aggregation’ method. Advantage of the method, that it makes possible the combination of working with lists of addresses from the Population Registry with reaching out to smaller populations not included in or not sufficiently covered by the registry data which are more difficult to find.
Any problems encountered in design and actions taken
Not all interviewers whom the subcontractors usually work with were comfortable with the new method applied, but the majority adjusted to the new requirements.
Description of sampling frame
When designing the sampling frame our aim was to represent the immigrant (foreign born) population in Budapest by their countries of origin to the highest possible extent. We identified 35 countries that represent 90 % of the foreign born population in Budapest.





Country

%

1

UA

26,4

2

RS

21,5

3

CN

14,6

4

VN

7,9

5

TR

3,2

6

US

2,8

7

JP

2,6

8

IR

2,3

9

MN

2,2

10

KR

1,8

11

IN

1,7

12

EG

1,6

13

IL

1,6

14

AF

1,5

15

SY

1,2

16

IQ

1,0

17

NG

0,9

18

RU

0,9

19

CA

0,5

20

HR

0,5

21

PS

0,5

22

YE

0,4

23

DZ

0,3

24

JO

0,3

25

LY

0,3

26

SO

0,2

27

AU

0,2

28

GE

0,2

29

LB

0,2

30

SA

0,2

31

SD

0,2

32

AM

0,1

33

KZ

0,1

34

MA

0,1

35

PK

0,1

We simplified the data collection by dealing with Arab countries as one unit due to their cultural and linguistic proximity. Based on similar considerations we merged the US, Canada and Australia into one group. When designing the sampling frame we used the following table. The table shows our plan (quota) and the actual sample composition (without the wights applied).




SAMPLE DESIGN: PLACE OF BIRTH

QUOTA (%)

SAMPLE

(%)

DIFF

(%)

ALL MUSLIM

13,4

13,3

-0,1

ASIAN "ENTREPRENEURS"

23,2

24,7

1,5

BORN IN EX JUGOSLAVIA

22,9

21,7

-1,1

BORN IN EX USSR

28,1

27,8

-0,3

WESTERN ENGLISH SPEAKER

3,7

3,8

0,1

Japan

2,6

2,6

0,0

Republic of Korea

1,7

1,8

0,1

Nigeria

1,0

0,8

-0,2

India

1,4

1,7

0,3

Israel

1,9

1,6

-0,3



DETAILED SAMPLE DESIGN

QUOTA (%)

SAMPLE

(%)

DIFF

(%)

ARAB COUNTRIES

6,7

6,3

-0,3

Turkey

2,9

3,2

0,3

Islamic Republic of Iran + Afghanistan

3,8

3,8

0,0

China

16,0

14,7

-1,3

Mongolia

1,8

2,2

0,4

Viet Nam

5,4

7,8

2,4

BORN IN EX JUGOSLAVIA

22,9

21,7

-1,1

BORN IN EX USSR

28,1

27,8

-0,3

Australia

0,6

0,1

-0,5

United States

2,4

3,2

0,8

Canada

0,7

0,6

-0,2

Japan

2,6

2,6

0,0

Republic of Korea

1,7

1,8

0,1

Nigeria

1,0

0,8

-0,2

India

1,4

1,7

0,3

Israel

1,9

1,6

-0,3

When designing the sample and collecting the data, we aimed at making the sample representative by sex and age, based on official macro statistical data provided by the Central Statistical Office (SOPEMI) and the Office of Immigration and Nationality.


Any problems encountered in completing sampling frame and actions taken
Official statistical data were not available in breakdowns we needed for creating the sampling frame, so we had to apply some estimations based on simplifying assumptions.
As there had not been similar survey in Hungary for the previous years, we did not have any information ont he characteristics of the population born in the „Western English-speaking” countries. During the fieldwork we realised that this population was not present in Budapest to the extent it had seemed from the official statistics based on residence permits. A possible explanation is that after the political transition many former emigrants and their descentants returned to and registered in Hungary, bought real estate and then returned to their countries of origin. Because of this we had to correct the actual number of people born in these countries (US, CAN, AU) using expert estimation, and adjust the sampling frame accordingly.
Description of how fieldwork was done (including list of all the specific centres of aggregation for those countries)
We worked with registry data using the ’centres of aggregation’ method. The registry contained data on foreign-born naturalised, long-term resident and refugee population. Data on people with short-term residence permit was collected in other centres of aggregation. When defining the centres we aimed at covering all possible aspects of public life of immigrants in Budapest. The following centres were used int he research:





centre (%)

attendance (%)

relative weight

a. Registry

52,12

93,6

1,00

b.      Kindergarten

0,42

17,3

0,18

c.       Elementary and secondary school

1,42

26,9

0,29

d.      Higher education institute, student dorm

5,75

29,8

0,32

e.      Other training centre (ie. Language school)

1,17

22,6

0,24

f.        Religious centre (mosque, prayer room, pagoda, chrisitan church)

2,41

40,8

0,44

g.       Market, shopping centre

9,74

93

0,99

h.      Immigrant shops (ie. indian grocery store, chinese shop)

3,08

54,5

0,58

i.         Immigrant restaurants, fast-food stalls

6,74

64,4

0,69

j.        Clubs, bars, nightclubs

1,75

44,3

0,47

k.       cultural or othjer social events

3,75

55,5

0,59

l.         Traffic centres, railvay or coach stations, travel agencies

2,25

73,9

0,79

m.    Offices of assisting agencies (ie. Menedék)

1,58

10,8

0,12

n.      Embassies, consulates

0,58

29,6

0,32

o.      Premises of immigrant community organisations

1,17

28

0,30

Public areas (open air)

6,08

94,4

1,01

The fieldwork was done by immigrants, university students speaking rare languages, students of sociology, andf professional interviewers alike. At the beginning there was a training/preparation session, and we operated a help-desk service during the fieldwork.


 Any problems encountered in fieldwork and actions taken
Due to administrative problems (the permit granted first lost int he post and we got it again only a very late stage of the fieldwork) we didn’t used the Office of immigration and nationality as a centre. We believe it is an important one, and could have helped us a lot in the work.
Using the registry data (list of addresses) proved considerably worse in case of the Chinese, Vietnamese, and Muslim sub-sample. Though we expected this, it needed additional coordination effort to handle this situation.

Description of your sampling methodology weightings
Weighting was done according to the methodology set in the paper: ’Centre sampling technique in foreign migration surveys: a methological note’ by Baio, Blangiardo, Blangiardo, consulting some details with Giancarlo Blangiardo.
Any problems encountered in designing the weightings and actions taken
As it was our first time using the ’centres of aggregation’ method, we did not have adequate information ont he frequency of visiting the centres we used. Thus the relative weight of each centre was established by expert estimations. Based ont he actual results, these estimations were not always well founded. Taking this into consideration we still believe that the data are not biased.
Your calculations of non-response rates and composition of reasons for non-response (based on contact form)
All questionnaires were annexed with a datasheet detailing the circumstances of the interview and the selection/finding of the respondent. These were only processed to the extent it was relevant for having the necessary information for establishing the weight variables. The ’centres of aggregation’ method does not require statistical correction handling response rates in the way it is required in case of random address list sampling.
List of specific questions where there may be problems of comparability with other countries (please specify which and why)

14. Do you want to become a citizen of Hungary?

The question does not deal with the issue/possibility of dual citizenship. Thus for example Chinese had a higher rate of saying ’no’ than in our previous surveys when the possibility of hypothetical dual citizenship was offered. Our experience is that one of the strongest determinants of immigrants’ attitude towards naturalisation is their country of origin’s policy toward dual citizenship. This factor can only indirectly be analysed in this dataset.


39. Have you completed a Hungarian language course provided OR funded by government in Hungary?

The question is restricting as it asks only about the state-funded courses. Respondents were often puzzled, as they were not 100% sure whether the course they took was state-funded or not. (Ie language training provided by an NGO, preparatory course provided by the university or college before enrolment etc)



58. Would you say that we need more MPs with an immigrant background in the Hungarian Parliament?

The question caused some interpretation problems, as due to the composition of the immigrant population there are (in fact quite a few) foreign-born MPs, but all of them are ethnic Hungarians from the neighbouring countries with no apparent immigrant identity markers. As there is no minority quota applied and there has never been any public discourse on thias issue, many of the respondents simply did not understand the concept and content of the claim ’having MPs with an immigrant background’.


Share with your friends:




The database is protected by copyright ©essaydocs.org 2020
send message

    Main page