18
1 Linguistic issues in survey research - The World Poll experience

Linguistic issues in survey research - The World Poll experience

Embed Size (px)

DESCRIPTION

 

Citation preview

Page 1: Linguistic issues in survey research - The World Poll experience

1

Linguistic issues in survey research - The World Poll experience

Page 2: Linguistic issues in survey research - The World Poll experience

This document contains proprietary research, copyrighted materials, and literary property of Gallup, Inc. It is for the guidance of your company only and is not to be copied, quoted, published, or divulged to others outside of your organization. Gallup® and The Gallup Poll® are trademarks of Gallup, Inc. All other trademarks are the property of their respective owners.

This document is of great value to both your organization and Gallup, Inc. Accordingly, international and domestic laws and penalties guaranteeing patent, copyright, trademark, and trade secret protection protect the ideas, concepts, and recommendations related within this document.

No changes may be made to this document without the express written permission of Gallup, Inc.

2

COPYRIGHT STANDARDS

Copyright © 2013 Gallup, Inc. All rights reserved.

Page 3: Linguistic issues in survey research - The World Poll experience

World Poll

Doing translations in survey research

Languages in sub-Saharan Africa

Challenges and opportunities

3

OVERVIEW

Page 4: Linguistic issues in survey research - The World Poll experience

Nationally representative samples of population aged 15 and older: 1,000 interviews per country per year

Pure, probability-based selection of:– Interviewing locations (Primary Sample Units)– Households– Adult respondent in household (regardless of age/gender)

Coverage: 120-140 countries per year, both urban and rural areas

Face-to-face, hour-long interviews in respondent’s home in developing world

Random-digit-dial telephone surveys (20 minutes) in developed world

Margin of error: +/- 3%

4

WORLD POLL – LARGEST INDEPENDENT SOURCE OF SURVEY RESEARCH

Copyright © 2013 Gallup, Inc. All rights reserved.

Page 5: Linguistic issues in survey research - The World Poll experience

5

Wellbeing

Law and Order

Governance

Jobs/Employment

Economics

Community Attachment

Migration

WHAT WE ASK

Entrepreneurship

Food and Shelter

Infrastructure

Health

Social Networks

Civic Engagement

Country Stability

Copyright © 2013 Gallup, Inc. All rights reserved.

Page 6: Linguistic issues in survey research - The World Poll experience

Keep questions simple

In the city or area where you live, do you have confidence in the local police force, or not?

Thinking about the job situation in the city or area where you live today, would you say that it is now a good time or a bad time to find a job?

Most World Poll questions are dichotomous – yes/no, satisfied/dissatisfied

6

CRAFTING SURVEY QUESTIONS

Copyright © 2013 Gallup, Inc. All rights reserved.

Page 7: Linguistic issues in survey research - The World Poll experience

Questionnaire

development

Translation

Pretest

Fieldwork

Data processing

Sampling

7

WORLD POLL PROCESS

Page 8: Linguistic issues in survey research - The World Poll experience

Preparation of core language questionnaires: English, French, Spanish, Arabic

Translations into all other World Poll languages – about 200

Independent review of translations: – 2 independent translations and a third party resolves

differences, OR– 1 person translates into target language, independent translator

back-translates to the source language, and a third party reviews/revises the translation

Language versions approved – Interviewers’ training starts followed by fieldwork

Interviewers must read the questionnaire word for word, spontaneous translations are not allowed

8

TRANSLATIONS

Copyright © 2013 Gallup, Inc. All rights reserved.

Page 9: Linguistic issues in survey research - The World Poll experience

COVERAGE – SUB-SAHARAN AFRICAN COUNTRIES

2013: 37 out of 56 African countries 95% of the population aged 15 and older

Since 2005-2006: Fielded at least one survey in every single country on the continent, except in: Cape Verde, Equatorial Guinea, Eritrea, Gambia, Guinea- Bissau, São Tomé & Príncipe, Seychelles, South Sudan

9

Page 10: Linguistic issues in survey research - The World Poll experience

COVERAGE – LANGUAGES IN SUB-SAHARAN AFRICA

Estimated number of African languages: 2035 (Grimes, 1996), or about 1/3 of all languages spoken around the world

Most are subdivided into dialects that are not mutually intelligible

In most countries, the linguistic diversity is very high:

Gabon - 40 languages - 1.5 million people – no real lingua franca DRC - 240 languages - 60 million people – 4 national languages

Most are unwritten languages National borders seldom correspond to linguistic

boundaries 10

Page 11: Linguistic issues in survey research - The World Poll experience

11

LANGUAGES IN SUB-SAHARAN AFRICA

High linguistic diversity

Most Africans speak at least two languages

Language a person uses depends on the type of information to be communicated and to whom it is communicated

Social stratification of languages:

Official Vehicular/Regional Mother tongue/Local

Page 12: Linguistic issues in survey research - The World Poll experience

AFRICAN LANGUAGES IN THE WORLD POLL CONTEXT

How to determine which languages to cover:

Recommendations from our local partners Other secondary research: Ethnologue.com database,

African Academy of Languages, etc. World Poll data about linguistic barriers

World Poll covers more 80 African languages (phonetic rendition)

12

Page 13: Linguistic issues in survey research - The World Poll experience

13

LANGUAGE OF INTERVIEW - DEMOGRAPHICS

SENEGAL KENYA

French Wolof English Swahili

Men 60 40 68 32

Women 53 47 62 38

15-29 65 35 74 26

30-45 53 47 67 33

46 and older 42 58 46 55

Primary complete or less 42 58 44 56

Secondary or higher 91 9 87 14

Urban 75 25 92 8

Rural 48 52 59 42

Page 14: Linguistic issues in survey research - The World Poll experience

14

LANGUAGE OF INTERVIEW - DEMOGRAPHICSBURKINA FASO ZAMBIA

French Moore

Dioula

English

Bemba Nyanja

Tonga Lozi

Men 68 32 1 39 26 22 11 3

Women 68 31 1 34 28 24 11 4

15-29 75 24 1 40 25 17 13 4

30-45 63 37 1 32 28 29 9 2

46 and older

59 40 2 30 30 28 8 4

Primary complete or less

87 13 _ 18 36 29 13 4

Secondary or higher

91 9 _ 53 18 17 9 3

Urban 79 20 1 42 33 21 4 --

Rural 66 33 1 34 24 24 14 4

Page 15: Linguistic issues in survey research - The World Poll experience

English/French is the language used most often to field the questionnaire

No gender differences

Younger respondents more likely to answer in English/French

More educated respondents are far more likely to use English/French

Source: 7 World Poll datasets from 2013

15

WORLD POLL LANGUAGE OF INTERVIEW - SUMMARY

Page 16: Linguistic issues in survey research - The World Poll experience

Challenges Identification of

appropriate languages Translation quality at

competitive prices Local staff capacity to

work with languages chosen for the project

Respondents’ social desirability to speak a language other than the official language

Actual respondents’ linguistic ability

16

LANGUAGES AS TOOLS TO MEET CUSTOMERS’ NEEDS

Opportunities Build knowledge about

the country/region of interest

Use local partners to help ‘localize’ the research project

Develop standards/best practices against which success can be measured

Page 17: Linguistic issues in survey research - The World Poll experience

17

English, of course! What

about you, young man?I didn’t

know Mama could speak

English!!

What language do you feel most comfortable speaking?

Page 18: Linguistic issues in survey research - The World Poll experience

18

DISCUSSION