DTM Surveys
Overview
Description
Surveys are a systematic way of asking people to volunteer information about their attitudes, behaviours, opinions and beliefs. Surveys target a specific individual who forms part of a group of interest to gather specific but non-identifiable information. Such information can include age, sex, motives, area of origin, transit points on the journey so far, intended destination, cost of the journey, and dangers encountered en route, etc. The success of survey research rests on how closely the answers that people give to survey questions matches reality that is, how people really think and act. Surveys can be used for descriptive research, causal explanations, evaluations of the efficacy of a program or policy, and to predict future events. Surveys can be used to verify information gathered through assessments, and to capture more information on specific topics that assessments cannot capture in sufficient detail.
Key differences between assessments and surveys are captured in the table below:
| DTM Assessment | DTM Survey |
| Defines problem | Explains problem |
| Usually the unit of reference is site/location | Unit of reference is household/individual |
| Rought estimates, not always accurate | Carefully selected sample with probability to represent whole population |
| Rapid and brief, no need for big resources | Slower, more detailed, more expensive |
| Quick overview for operational use | Provide smore in-depth understanding to support planning, design and develop projects, improve operational response |
| Key informant interviews and direct observation |
Face-to-face interviews Online/phone/mail surveys |
A sample is a smaller, hopefully representative collection of units from a population, used to determine truths about that population selected. Using a sample enables to receive results with known accuracy that can be calculated mathematically. The sampling frame is the list from which the potential respondents are drawn. The sampling unit can be at individual or household level. The appropriate sampling method varies depending on the context and includes probability and non-probability methods. Probability sampling is the best approach if you want to generalize to the populations as a whole, compare random events in the results. Use closed-ended questions in surveys with quantitative sampling. Non-probability sampling is recommended in cases where probability sampling is considered inefficient and does not provide the information of interest. Use open-ended questions, for example through focus group discussions, in surveys with qualitative sampling.
| Method | Description | |
| Probability/Quantitative | Simple random | Get a list or "sampling frame" that must not systematically exclude anyone. Generate random numbers. Select one person per random number. |
| Systematic random | Select a random number, which will be known ask. Get a list of people (sampling frame). Select every k-th person. | |
| Stratified | Separate your population into groups or "strata". Do either a simple random sample or systematic random sample from there. | |
| Multi-stage cluster | Get a list of "clusters," e.g., segments within an area. Randomly sample clusters from that list. Have a list of, say, 10 branches. Randomly sample people within those branches. | |
| Non-probability/Qualitative | Convenience | Find some people that are easy to find. |
| Snowball | Find a few people that are relevant to your topic. Ask them to refer you to more of them. | |
| Quota | Determine what the population looks like in terms of specific qualities. Create "quotas" based on those qualities. Select people for each quota. |
The sample size has an impact on the reliability of the survey results. A sample size calculator is used to determine the required sample size to achieve a desired confidence rate (e.g. 95% or 99%).
Good questions are clear and use simple language; concise; specific; possible to answer; relevant to the respondent. Furthermore, they do not use negatives; avoid bias terms as well as leading questions; and have only one part.
The questionnaire format should be as short as possible, visually attractive and readable (consider font size), uncluttered and numbered, broken into logical sections if possible, feature clear skip patterns for contingency questions and have clear spaces for respondents to mark answers, using boxes, parentheses (X), or numbers to circle. Avoid lines to put check___.
Examples from the field
The locations where flow monitoring surveys were conducted in the Western Balkans and Mediterranean were commonly located at both official and unofficial entry, transit and/or exit points. In line with fluctuations in scope and directions of flows in the Mediterranean crisis, it was often required to adjust the locations where the surveys are conducted. For example, after 8 March 2022, when the Western Balkans route was closed, the survey focus shifted to the migrant population that remained stranded in Western Balkan countries, with interviews held in official camps in Northern and Southern Greece. The option to interchangeably use paper and electronic forms to record survey response allowed operational flexibility and facilitated data collection. Additionally, the existence of the forms in variety of languages spoken by the migrants along the route allowed for data collection even when the communication in English or French (languages most commonly spoken by the data collectors) or the availability of interpreters was difficult/challenging.
Contacts
For more information, please contact the DTM Support Team: [email protected].