Consumer research solutions
WorldSurveys® market
measures
Background
WorldSurveys provide market assessments or other usage and attitude
measures in
emerging cities or countries as well as more developed ones.
- In emerging cities or countries, research infrastructure very often either does not exist, or leaves much to be desired in terms of coverage and quality. 1WR's approach recognizes the inadequacy of much rural research - and in the more difficult parts of many urban areas - and is committed to provide a more comprehensive and rigorous approach to market research.
- In developed areas, 1WR provides expertise in recruiting, training and managing field teams to ensure professional standards of data provision.
Objectives
To draw a sound representative sample that will stand up to the most
critical analysis and provide good estimates of:- the total numbers of buyers, users or subscribers: population size, described by key demographics, for categories, brands, distribution channels
- usage and habits
- attitudes and imagery
- eventually some diagnostic information to facilitate brand positionning or other marketing improvements
Sampling
By offering samples drawn to represent rural as well as urban areas, as
comprehensively as is possible and practicable, WorldSurveys research
offers a genuine addition to research services currently on offer.
The general sampling objectives are:
- to generate samples which conform to the best practice in survey research
- to adapt such best practice to the needs and circumstances of the area or territories being surveyed
- to achieve the most comprehensive coverage of both urban and/or rural areas commensurate with client's demand
- each sampling plan should be as robust and as transparent as possible, and exhibit the maximum achievable degree of randomness
- samples should ideally be designed to be representative of individuals
- quotas should be avoided wherever possible, partly since the information on which to base them may be lacking, and partly since quotas can introduce elements of interviewer bias
- the universe should be regarded as adults, generally aged 15 and over
- sampling plans are drawn up after consultation with local statistics offices, census authorities and university departments, as appropriate
- local statistical data, whether census data or other material, are validated wherever possible against other sources (for instance, UN agencies, Population Reference Bureau...)
- areas such as war zones, parts of a country with high levels of lawlessness, and areas where the practical difficulties - and hence costs - of reaching certain remote populations will be clearly delineated in any sampling plan
Stratification, sampling points,
selection of respondent
Given the frequent lack of comprehensive sampling frames in many
countries,
some form of multi-stage stratified random sample is employed. The
criteria for stratification may vary from territory to territory.
Depending on the nature of the representativeness required, the first
stage would be stratification by
province or by district clusters.
Additional stratification criteria and stages would depend on the territory being sampled, but might typically include community size, district type, ethnicity/tribal affiliation, gender and age group.
Sampling points are selected, at random and proportional to population size and distribution, from such lists of cities, districts, towns and villages as may exist and be accessible.
The selection will allow for alternate sampling points, should difficulties be encountered with the initial selection.
There is a limit on the number of interviews held at each sampling point: a maximum of 10 is recommended.
Individual households within each sampling point should be chosen by a random route method.
A fieldwork starting point should be selected at random from available documentation or observation, from which the random walk should commence. In multi-storey dwellings, criteria should be established for household selection.
Once a household has been selected, only one person from that household is interviewed. The method of selection for the interviewee should permit the maximum degree of randomness. Substitution within the selected household should not be permitted. If it has not proved possible to interview the selected person after an agreed number of call-backs, then another household should be substituted. Three attempts are suggested as the norm.
Interviews are conducted face to face, in home.
It may be the case, particularly in remote rural areas, that significant numbers of selected interviewees are always away from home when interviewers call during normal hours, and that call-backs are not practical due to the remoteness of the sampling point. In such cases, interviewing at their place of work is be considered.
Cultural and local considerations
Surveys are scheduled when possible to avoid seasonal bad weather or
religious
holidays.Other survey design factors:
- each survey will offer typical baseline demographic questions, general media usage and attitude questions, and more specific media questions
- the interview' total length should not exceed 35 minutes, unless incentives are provided
- questions will wherever possible be closed, with a maximum of 5 open-ended questions per interview
- all questions to be carried on the survey will be piloted in each territory where the survey is run
- particular attention is paid to translation issues and language of interview
- questionnaires are prepared in the major languages of the territory in question, and every effort will be made to have interviewers on the team who would be able to conduct an interview in a local language for which no printed translation is available
This is an example of a successfully completed project. Specific parts
of the project are retained to protect confidentiality.
Difficulty:No established national research services were available.
Solution: 1WR Project Director visited a number of diverse organisations with potential national field workers. Having reviewed their team leaders’ experience, a consultancy was selected having supervisors speaking all local languages.
Door to door; multi-stage random selection of 1000 households; 30 minute questionnaire on photo usage and attitudes.
Difficulty: Field workers needed to work in remote areas where supervisors might not always be available, leading to poor control of teams in remote areas.
Solution: Field workers underwent intensive training with supervisory follow-up in the main city. A separate quality team was established under the Project Director’s authority independent of the field teams to randomly check fieldwork in remote areas.
The client developed sound marketing activities from this quality research.
Controls consisted of back checking a number of completed interviews.
Objectives
To assess the size of the photo market in an emerging market: Uganda .Difficulty:No established national research services were available.
Solution: 1WR Project Director visited a number of diverse organisations with potential national field workers. Having reviewed their team leaders’ experience, a consultancy was selected having supervisors speaking all local languages.
Sampling and methodology
MethodologyDoor to door; multi-stage random selection of 1000 households; 30 minute questionnaire on photo usage and attitudes.
Difficulty: Field workers needed to work in remote areas where supervisors might not always be available, leading to poor control of teams in remote areas.
Solution: Field workers underwent intensive training with supervisory follow-up in the main city. A separate quality team was established under the Project Director’s authority independent of the field teams to randomly check fieldwork in remote areas.
Reporting and advice
National estimates and market characteristics were detailed in a report
and summarised in a slide presentation. A field appendix indicated all
field details as well as training and controls.The client developed sound marketing activities from this quality research.
Project duration
Project duration: 6 weeksQuality
The project manager formed a separate quality team who followed the
field team
during the whole operation.Controls consisted of back checking a number of completed interviews.
© Copyright 1 World Research (1WR.net)
2002-2008. All Rights Reserved.
updated last Sept. 2004
updated last Sept. 2004