Understanding Social Research Methods

Social Research Methodsin the study of ecosystem services

Mahesh Poudyal, P4ges Project

Bangor University

Ecosystem Services for Poverty Alleviation

Summer School

13 April 2016

13/04/2016Social Research Methods, ESPA Summer School 20162

Outline


‘Social’ research in the study of ecosystem services

Evolution of social and participatory research methods

Collecting (social science) data

Key issues and challenges in social and household surveys

Why do we need ‘social’ research methods in

the study of ecosystem services?


Conceptual Framework of the

Millennium Ecosystem

Assessment 2005

(MA, 2003)




Linkages between

ecosystem services

and human well-being

(MA, 2005)




“Ecosystem services are the benefits provided by

ecosystems that contribute to making human life

both possible and worth living”- UK National Ecosystem Assessment

Better understanding of the ‘human dimension’ crucial

in studying the changes in ecosystem services, and how

such changes affect human well-being

Good data on human-ecosystem services-poverty

nexus still limited

Social and participatory research


Participatory research

Origins, evolution, benefits, critiques

Participatory methods

Community mapping, timeline & trends, seasonal calendar etc.

Mixed method approach in collecting socioeconomic data

Qualitative & quantitative surveys, participatory approaches

Key issues & challenges

Sampling, project presentation, consent, confidentiality, feedback

Participatory Research


Origins of participatory research


Rapid Rural Appraisal

1970s-80s

Participatory Rural

Appraisal

1980s-90s

Participatory Action

Research

1980s-90s

New approach to rapid data-gathering, using informal & locally appropriate techniques

More emphasis on participation – respondents drive research questions and collect data themselves

No longer researchers and respondents – people are learning together and then acting on that knowledgeEm

po

wer

men

t

Res

earc

he

r b

eco

mes

a li

sten

er

(Chambers 1994)

RRA & PRA


(Chambers, 1994)

Principles of participatory methods


The right to participate

Seek local knowledge and experience

Seek diversity

Create safe spaces for people to be heard

Triangulate

What benefits for the locals?


Make it clear to the local people what they will gain

by being involved in the research

Then start building trust and a common

understanding of what will be achieved

Critiques of the participatory approach


"Hijacking“ - The agenda is externally driven, and

used to create legitimacy for projects, agencies and

NGOs

Formalism - The “hit team" arrives in a local

community to "do a PRA"

Disappointment - Local expectations can easily be

raised

Participatory approaches


Community mapping (social, resource)

Transect walks

Timeline

Trends

Seasonal calendar

Wealth ranking

Identification of stakeholders & interests

Observations

Key Informants

Focus Groups

Community mapping


Good ice-breaker

Can get different

members of the

community to do it

(Photo: P. Ranjatson)

Focus groups


Focus groups are guided discussions that are often recorded

Qualitative data can then be extracted

It’s a discussion not a debate; important to allow people to

voice opinions

Need a clear set of questions and aim

Identify who are the most appropriate participants

Can be used to help develop

questionnaires and further study

Focus groups


Outcome of a community mapping exercise

in Madagascar for p4ges project.


Timeline & Trends


Best done in a focus group with village elders

How has the forest extent and quality changed in this time, and why?

Seasonal calendar


Don’t make it too complicated! Bean counts are useful

What are the main seasons of the year?Can you tell me what men and women are doing in each season?

Wealth ranking


Useful for: understanding concepts of wealth, stratification,

getting a single metric for wealth for quantitative analysis

Pitfalls: can be sensitive topic, people need to know each

other well, diverse meanings of wealth

Summary of techniques


Ranking (e.g. taste preferences, most hunted species)

Drawing (e.g. maps)

Discussing (focus groups)

Apportioning (bean counts)

Picture recognition (e.g. which species live round here)

Be inventive and use locally appropriate

techniques and materials

Problems with participatory research


Ignoring marginalised groups (women, low caste, geographically

isolated, children)

Focus groups may be dominated by certain individuals and

“group think” may emerge

Leads to unrepresentative results

Some cultures find it patronising, some threatening

Can be hard to generalise and analyse from qualitative results

– may be better as an early component of a broader study

Collecting social science data:

a mixed method approach13/04/2016Social Research Methods, ESPA Summer School 201624

Social research methods


Quantitative Qualitative

Questionnaire surveys Participatory research

MIXED

Quantitative method


Variable-centred approach in data collection

Can collect large quantities of data

Often used to test theories/hypotheses

Aim is to be representative of the ‘population’

Relatively less expensive

Questionnaire surveys


Survey design

DO NOT start from scratch, there are numerous surveys

available which could be adapted to your needs

For example, PEN prototype surveys

Or the World Bank’s LSMS

Characteristics of ‘good’ questions (Angelsen et al. 2011)

KISS (keep it

sensibly simple)

Concrete

and specificShort

recall

Local

units

Quantify

answers

Define terms

carefully

Avoid multipart

questionsNeutral

formulations



Pretesting/piloting questionnaire surveys crucial

Improving questionnaire

Training enumerators

Establish field protocols for data collection

Include ethical framework

Practical issues; e.g., dealing with dropouts from sample

Prepare detailed guidelines for survey

A section from the World Bank’s Living Standard Measurement Survey (LSMS) in Nepal



1

Work Package 6: Household Questionnaire Survey FID: ________ HHID: ________

B. Information about the Respondent(s) and Household Composition

1. Information about the household head and the respondent

a. Is the main respondent also the head of the household? [If 'YES' >> go to 'c.'] (YES=1, NO=0)

b. What is the relation of the respondent to the HH head? [use code from 2 below] (code)

Codes for relation to HH head: 1=spouse(legally married or co-habiting); 2=son/daughter; 3=son/daughter in-law; 4=grandchild; 5=mother/father; 6=mother/father in-law;

7=brother/sister; 8=brother/sister in-law; 9=uncle/aunt; 10=nephew/neice; 11=step/foster child; 12=other family; 13=not related

[Note: Once you have established whether the respondent is also the household head from above, ask the following questions to the respondent and record their

responses in the appropriate column. When the respondent is NOT the household head, also ask the respondent about the household head and fill in the respective

column (for c. to i.).]

Respondent IS the HH Head Respondent is NOT the HH Head

c. Was the respondent born in this village? [If 'YES' >> go to 'g.'] (YES=1, NO=0) (YES=1, NO=0)

d. How long has the respondent lived in the village? _______ years _______ years

e. Where did the respondent come from (commune and region)?

f. Why did the respondent move to this village? (code) (code)

1=study; 2=work; 3=marriage; 4=land availability; 5=other (specify)

g. Which ethnic group does the respondent belong to? (code) (code)

Ethinc group: 1=Betsimisaraka; 2=Sihanaka; 3=Bezanozano; 4=Merina; 5=Betsileo; 6=Other (specify)

h. What is the marital status of the respondent? (code) (code)

1=married & living together; 2=married & spouse working away; 3=widow/widower; 4=divorced; 5=never married; 6=other (specify)

i. How long ago was this HH formed? ________ years ________ years

[(Non-)response codes: -98=respondent DOES NOT KNOW; -99=respondent DOES NOT WANT TO ANSWER; -100=interviewer DID NOT ASK (specify why)]

Remarks:



WP6SurveyGuidelines,Draft1.0,July2014

1

P4GESWorkPackage6:GuidelinesforSurvey

KeyDefinitions

Households [also need to look into Madagascar-specific definition, i.e., how does INSTAT define a

household in Madagascar?]

Working definition: A household is defined as one person living alone or a group of people living

together, who pool some, or all, of their resources (labour, income and wealth) and who make common provision

for food or other essentials for living. It is important to note that it is possible to have household members who

are no blood relatives of the family. Equally important to note is that one single household might include more

than one family if they pool their resources and eat together as defined above.

Polygamy: In the case of polygamous families where each wife lives separately, the husband is

considered part of the household where he normally resides and shares food - with each separately-living wife

constituting a household each.

Multiple families living together in one house: If there are more than one family living in one house, they

should be treated as two separate households if there is no resource pooling and they mostly cook and eat

separately.

Family members who partially live away (for example, for school or for work): These members will still

be part of the household if they are more or less fully integrated in the household economy (for example, school

children living away during the week with parents paying for their expenses; or members working away part of

the time but pooling their income, and pooling their labour in the households rest of the time).

Non-resident household members: Among the household members who partially live away (see above),

if they live away for six months or more in a year then they should be identified/indicated as ‘non-resident’

members of the household.

Family members who live away on their own: For the family members who have left the house to work

and take care of themselves but still contribute some income to the remaining household, they should NOT be

included in the household as members BUT the income they contribute to the household should be recorded

under ‘remittances’.

Household head: This should be the person who makes most of the major decisions within the

household, and is recognised as such by the local authorities (customary and/or statutory). In most case this

would be the husband of the family, however, this might be more complicated if several generations or multiple

families are living together in the same household. In such situation, asking “who is the head of the household?”

would generally help identify the head.



Different ways of administering questionnaires

Face to face

Self-administered

Telephone interview

Internet surveys

Face to face surveys most effective and practical in a

poor developing country context

Qualitative method


Meaning-centred approach – emphasis on words in

context rather than numbers

Less structured and flexible – can collect large volume of

information from few subjects

Used to create theories/hypothesis

Qualitative method


Aim is to describe the ‘issues’ as the subjects understand,

perceive, construct them

Skills for data collection often acquired through practice

Relatively more expensive

Collecting qualitative data: Participatory

approaches


Community mapping (social, resource)

Transect walks

Trends

Seasonal calendar

Timeline

Wealth ranking

Identification of stakeholders & interests

Observations

Key Informants

Focus Groups

Some ways of combining quantitative

and qualitative methods (Bryman 2006)


Instrument development: e.g., qualitative methods used to

develop questionnaires

Sampling: one method used to facilitate sampling for the

other

Utility (improving the usefulness of findings): combining

two methods likely to be more useful to applied/action

research

Triangulation: verify/corroborate both data and findings

Explanation: one method used to explain findings from

the other

Which method to use?


Qualitative (participatory) approaches are generally

better as a foundation for conservation action

Quantitative approaches may produce more defensible

research results

Mixed methods, if applied well, could provide the best of

both approaches


The approach you choose depends

on your research aims, your

budget and the social context of

your study site.

ClassCaste

GenderEthnicity

CustomsIndigenous

MigrantsRural

Urban

Challenges in social research &

how to overcome them


Sampling


Need a robust approach to sampling so you can generalise from

the results

Unlikely to get complete

census so we take a sample

from a defined population

How do we know if the

sample is representative?

Sampling strategies


Random sampling

Develop a sampling frame and use random number generator

Stratified – random sampling from pre-defined strata

Systematic sampling

Every 5th person, every 50 metre… Straightforward and relatively

objective (still an unknown potential for bias)

Sampling strategies


Opportunistic sampling

Get data from anyone you can; often used e.g. for tourist samples. Can

get you big samples quickly

Purposive sampling

Choose who you want to interview; includes snowball sampling (ask

interviewees to recommend the next person)

Good if few people are doing what you’re interested in, or introductions

are helpful

Good for qualitative research

Sampling strategies


Sampling approaches have to balance theoretical & practical

considerations, and are often mixed in practice

For example:

purposive selection of sites, communities

stratification based on wealth ranking, or geographical location

random sampling from each stratum (Angelsen et al., 2011)


Selection of Study Sitesand Sampling for WP6

BENEFITSASSESSMENT

OPPORTUNITYCOST OF TAVY

Next to MNP (strictconservation)

Purposivelyselected villageswithin ONE orMORE fokontany(s)

Safeguards(conservation w/compensation)

MIXED communitiesaround CAZ

Purposively selected villagesfrom 5-6 fokontanys around CAZthat have had conservation-related projects in the past orhave ongoing conservation-related projects (such as Node,Conservation Agreements etc.)

No conservation (communitiesw/ historic deforestation -along a gradient)

Purposivelyselected villageswithin ONE orMORE fokontany(s)

Purposively selected villages within ONE or MOREfokontany(s) from each of the identified sites alongthe gradient: (1) forest edge communities withexisting deforestation & no recent history ofconservation; and (2) communities far from thepresent forest edge with history of complete forestclearance (within the last 20-30 years)

1. Communities around CAZ that are/likelyto be affected by the new protected areas

General Criteria forSite Selection

2. Security 3. Accessibility

Exact sites yetto be identified

(2) Far from forest: Purposively selectedvillages in fokontanyAmporoforo incommune Amporoforo

Purposively selectedvillages in fokontanyAmpahitra incommune Ambohibary

Purposively selectedvillages in fokontanysAntevibe & Ambodivoangyin communeAmbodimangavalo

Identified/Selected Sites Identified/Selected Site Identified/Selected Site

(1) Forest edge:Purposively selectedvillages in fokontanySahavazina incommune Antenina

Identified/Selected Sites

Study aims and specific criteriafor the selection of study sites

50-60 HHsin each siteincludingreplacement

Total #HHs inselected villagesin each site

Total #HHs inselected villages




175 HHs incl.replacement




Selected Sites* & Sampling Frame**

RANDOM RANDOM RANDOMRANDOM RANDOM

In-depth agriculture survey and in-depth wild product harvest survey: Stratified random selection

50 HHs 50 HHs 50 HHs 50 HHs

Sampling Method & Sample Size***

Final number of HHs to be surveyed

Notes:* Number of hamlets will depend upon their size and other relevant characteristics.** List of households in the selected hamlets will be collected from the relevant fokontany/commune and verified on site by the researchers with the help of local Key Informants.*** Number of HHs indicated as sample size includes roughly 10-15% additional households for replacements if some HHs drop out from the survey.**** Choice Experiment will not be done in “far from forest” gradient site.

Assessment of Benefits40-50 HHs in each site

150 HHs 150 HHs **** 200 HHs 150 HHsGeneral HH Survey &Choice Experiment

Framework for sites selection

& sampling in P4ges project

Purposive selection of sites

based on predefined criteria

Decision on sampling frame;

stratification by villages

(geographic)

Sample size requirement;

sampling of HHs for

survey (stratified random

sampling)

Sampling challenges in developing country

context


Poor existing data on

Population (distribution)

Location of communities (e.g., maps not updated)

Households scattered over large area; shift with season

Households not registered in any village (‘unofficial’ HHs)

How to deal with

individuals who want

to be interviewed?

Overcoming sampling challenges


Spend more time in the field!

In one of the P4ges project study sites, about 33% of

total time for survey…developing sampling frame

Fokontany level

Village level

Hamlet level

Collect locally available information on

villages (sketch map)

Collect information on households and

hamlets (sketch map & GPS)

Visited hamlets in person to cross check

information (GPS), and map HHs location

(sketch map & GPS)

46

Developing sampling frame: Sketch map of a village showing hamlets & households

(P4ges project)

Overcoming sampling challenges


Make the process transparent

Draw samples in front of community leaders, elders

Non-selection of key community members could

be an issue

Use them as key informants so they feel valued

If you have to include them in survey to ensure cooperation,

drop them from analysis

What sampling challenges have

you faced in your research?


Research ethics


How should we treat the people on whom we conduct

research?

Are there activities in which we should or should not

engage in our relations with them?

Research ethics


Ethics is about a collective value separating right from wrong

Research has to be conducted respecting ethics principles

agreed for the given context

Respect others and their autonomy

Protect those with diminished autonomy

Do good, not harm

Consider justice and equity

Act with integrity, honesty and safety

Be responsible towards others

Ethical dilemmas in research


You come across a lemur trap in the forest, would you

report it to the relevant authorities?

You are taking some soil samples in the first month of the

project and a villager asks you how the project is going to

benefit him. How would you respond?

Research ethics


A good ethics framework should

minimize misunderstandings and conflicts about or around the

project…

through a high level of participatory agreement that should be

reached from the beginning

Some fundamental principles that everybody needs to

consider

Informed consent


Free

Voluntary participation, no pressure, no condition

Prior

Before any data collection

Informed

Purpose of research/project, consequence of participation

Consent

Explicit (oral, written), temporary

Project presentation


Consistency in how you present the

project

Be clear on what the project is about

what it will/not provide to participants

whether it is purely a research project

whether it has any immediate tangible

benefits

Don’t create unrealistic expectations!

Confidentiality, data security


Balance the risk to respondents with the duty to publish

results

Data ownership – whose data?

Different layers of data security:

Anonymity

Secure storage

Proper data sharing protocol


Doing excellent research

with impacts

Key is to find a balance

Rights (and safety) of those

providing information


Need for appropriate and strict

Data Firewalls within the project

& beyond

…starts with a detailed ‘Data

Management Plan’


Some

identifier

information

necessary for

follow-up

surveys

Stripped of all

identifier

information

P4ges project data

management framework

Data sharing within projects & beyond


Data management plan

Identify and agree on the existing/potential conflict of

interests

Agree on a course of action, including establishment of

proper data firewalls even within the project

Remember…

Respondents have trusted you with their information,

it is your responsibility uphold that trust


Have you faced any challenge in data

management within your own research

project?

Restitution (reporting back)


We have a duty to communicate results of the project to

different stakeholders

Most importantly to the communities providing data

Booklets to communicate research results from P4ges to

local communities (final version will be in local language)

Directly sharing

the knowledge

generated

through

research is one

way of

benefiting local

communities

Compensation


“Time is money”

Agree on project-wide

compensation policy for

Local workers, facilitators

Key Informants, respondents

Communities

In-kind compensation may be

more appropriate

Some of the items in P4ges project

compensation package

Compensation


Could lead to people wanting to be interviewed for

compensation

Make compensation fair but not too attractive

How have you compensated social survey

respondents for their time?

What were the challenges in providing compensation?

Dealing with ethical dilemmas: some cases

to consider


What is the ethical issue at stake?

What would you do?

What alternatives might also be ethical?

What ethical issues remain with your chosen course of

action?

Scenario 1


You are working in a village and have made an acquaintance

with a poacher. He is the ideal resource person. Do you

have the right to request information? You ask him for

"informed consent" and he refuses. Do you think you

should ask him for any further information?

Scenario 2


You just learned this morning that the access to the forest,

which is ideal for your inventory today, is prohibited

according to the local custom. What would you do?

Scenario 3


You encounter a forest service officer in your field work

about illegal logging. And he asks you, "I heard that there are

illegal loggers in the area where you are coming from.

Where is this place exactly, and do you know how many

people are involved and where they come from?” – How

would you respond?

Scenario 4


You are camping in the forest for inventory. You come to

know that one of your workers (local assistant) established

lemurs traps during last night. What would you do?

Dealing with ethical dilemmas: some cases

to consider


What is the ethical issue at stake?

What would you do?

What alternatives might also be ethical?

What ethical issues remain with your chosen course of

action?


10 Minutes


What ethical challenge have you had

to deal with in your field research,

and how did you resolve it?

Summarising social research process


Before

fieldworkFieldwork

After

fieldwork

Plan

Design

Prepare

Collect

Analyse

/publish

Share

General & specific

considerations: ideas, research

questions, hypotheses

Frameworks, data needs,

sampling, contextual

information, survey design

Practical considerations:

logistics, obtaining permits,

learning language etc.

Political & cultural issues,

survey implementation, quality

checks

Data entry, quality checks,

cleaning, analysing, writing

reports, papers

Presenting results

(conferences, workshops),

sharing with community

(Adapted from Angelsen et al. 2011, p.10, Fig 1.3)

Key to successful field research


Stay in the field as much as possible

Have a dedicated team of enumerators

Building relations with respondents is the key to

success

Check and double-check data

Have a plan – and a plan B

(Angelsen et al., 2011)

Useful resources


Angelsen et al., 2011. Measuring Livelihoods and Environmental

Dependence: Methods for Research and Fieldwork, CIFOR

Poverty Environment Network (PEN) -

http://www1.cifor.org/pen

World Bank’s Living Standards Measurement Study -

http://go.worldbank.org/IFS9WG7EO0

Grosh & Glewwe, eds., 2000. Designing household survey

questionnaires for developing countries: lessons from 15 years

of the Living Standards Measurement Study, Vols. 1-3.

UN Statistics Division, 2005. Household Sample Surveys in

Developing and Transition Countries

UN Statistics Division, 2005. Designing Household Survey

Samples: Practical Guidelines

Bryman, A., 2015. Social Research Methods, 5th Edition

Krueger, R.A. (1988) Focus Groups: A practical guide for applied

research. Sage, UK.

Krishna Kumar (1986) Conducting Key Informant Interviews in

Developing Countries, AID Program Design and Evaluation

Methodology Report No. 13. December 1986.

http://pdf.usaid.gov/pdf_docs/PNAAX226.PDF

http://www1.cifor.org/pen

http://go.worldbank.org/IFS9WG7EO0

http://pdf.usaid.gov/pdf_docs/PNAAX226.PDF

Understanding Social Research Methods

Documents