Survey methodology


Survey methodology is "the study of survey methods".
As a field of applied statistics concentrating on human-research surveys, survey methodology studies the sampling of individual units from a population and associated techniques of survey data collection, such as questionnaire construction and methods for improving the number and accuracy of responses to surveys. Survey methodology targets instruments or procedures that ask one or more questions that may or may not be answered.
Researchers carry out statistical surveys with a view towards making statistical inferences about the population being studied; such inferences depend strongly on the survey questions used. Polls about public opinion, public-health surveys, market-research surveys, government surveys and censuses all exemplify quantitative research that uses survey methodology to answer questions about a population. Although censuses do not include a "sample", they do include other aspects of survey methodology, like questionnaires, interviewers, and non-response follow-up techniques. Surveys provide important information for all kinds of public-information and research fields, such as marketing research, psychology, transportation studies - the study of how people and goods move across space, health-care provision and sociology.

Overview

A single survey is made of at least a sample, a method of data collection and individual questions or items that become data that can be analyzed statistically. A single survey may focus on different types of topics such as preferences, opinions, behavior, or factual information, depending on its purpose. Since survey research is almost always based on a sample of the population, the success of the research is dependent on the representativeness of the sample with respect to a target population of interest to the researcher. That target population can range from the general population of a given country to specific groups of people within that country, to a membership list of a professional organization, or list of students enrolled in a school system . The persons replying to a survey are called respondents, and depending on the questions asked their answers may represent themselves as individuals, their households, employers, or other organization they represent.
Survey methodology as a scientific field seeks to identify principles about the sample design, data collection instruments, statistical adjustment of data, and data processing, and final data analysis that can create systematic and random survey errors. Survey errors are sometimes analyzed in connection with survey cost. Cost constraints are sometimes framed as improving quality within cost constraints, or alternatively, reducing costs for a fixed level of quality. Survey methodology is both a scientific field and a profession, meaning that some professionals in the field focus on survey errors empirically and others design surveys to reduce them. For survey designers, the task involves making a large set of decisions about thousands of individual features of a survey in order to improve it.
The most important methodological challenges of a survey methodologist include making decisions on how to:
  • Identify and select potential sample members.
  • Contact sampled individuals and collect data from those who are hard to reach
  • Evaluate and test questions.
  • Select the mode for posing questions and collecting responses.
  • Train and supervise interviewers.
  • Check data files for accuracy and internal consistency.
  • Adjust survey estimates to correct for identified errors.
  • Complement survey data with new data sources

    Selecting samples

The sample is chosen from a sampling frame, typically a list of population units, although in area sampling the frame may consist of a map that delineates geographic units.
The goal of a survey is not to describe the sample, but the larger population. This generalizing ability is dependent on the representativeness of the sample, as stated above. Each member of the population is termed an element. There are frequent difficulties one encounters while choosing a representative sample. One common error that results is selection bias. Selection bias results when the procedures used to select a sample result in over representation or under representation of some significant aspect of the population. For instance, if the population of interest consists of 75% females, and 25% males, and the sample consists of 40% females and 60% males, females are under represented while males are overrepresented. In order to minimize selection biases, stratified random sampling is often used. This is when the population is divided into sub-populations called strata, and random samples are drawn from each of the strata, or elements are drawn for the sample on a proportional basis.

Modes of data collection

There are several ways of administering a survey. The choice between administration modes is influenced by several factors, including
  1. costs,
  2. coverage of the target population,
  3. flexibility of asking questions,
  4. respondents' willingness to participate and
  5. response accuracy.
Different methods create mode effects that change how respondents answer, and different methods have different advantages. The most common modes of administration can be summarized as:
  • Telephone
  • Mail
  • Online surveys
  • Mobile surveys
  • Personal in-home surveys
  • Personal mall or street intercept survey
  • Mixed modes

    Research designs

There are several different designs, or overall structures, that can be used in survey research. The three general types are cross-sectional, successive independent samples, and longitudinal studies.

Cross-sectional studies

In cross-sectional studies, a sample is drawn from the relevant population and studied once. A cross-sectional study describes characteristics of that population at one time, but cannot give any insight as to the causes of population characteristics because it is a predictive, correlational design.

Successive independent samples studies

A successive independent samples design draws multiple random samples from a population at one or more times. This design can study changes within a population, but not changes within individuals because the same individuals are not surveyed more than once. Such studies cannot, therefore, identify the causes of change over time necessarily. For successive independent samples designs to be effective, the samples must be drawn from the same population, and must be equally representative of it. If the samples are not comparable, the changes between samples may be due to demographic characteristics rather than time. In addition, the questions must be asked in the same way so that responses can be compared directly.

Longitudinal studies

Longitudinal studies take measure of the same random sample at multiple time points. Unlike with a successive independent samples design, this design measures the differences in individual participants' responses over time. This means that a researcher can potentially assess the reasons for response changes by assessing the differences in respondents' experiences. Longitudinal studies are the easiest way to assess the effect of a naturally occurring event, such as divorce that cannot be tested experimentally.
However, longitudinal studies are both expensive and difficult to do. It is harder to find a sample that will commit to a months- or years-long study than a 15-minute interview, and participants frequently leave the study before the final assessment. In addition, such studies sometimes require data collection to be confidential or anonymous, which creates additional difficulty in linking participants' responses over time. One potential solution is the use of a self-generated identification code. These codes usually are created from elements like 'month of birth' and 'first letter of the mother's middle name.' Some recent anonymous SGIC approaches have also attempted to minimize use of personalized data even further, instead using questions like 'name of your first pet. Depending on the approach used, the ability to match some portion of the sample can be lost.
In addition, the overall attrition of participants is not random, so samples can become less representative with successive assessments. To account for this, a researcher can compare the respondents who left the survey to those that did not, to see if they are statistically different populations. Respondents may also try to be self-consistent in spite of changes to survey answers.

Questionnaires

Questionnaires are the most commonly used tool in survey research. However, the results of a particular survey are worthless if the questionnaire is written inadequately. Questionnaires should produce valid and reliable demographic variable measures and should yield valid and reliable individual disparities that self-report scales generate.

Questionnaires as tools

A variable category that is often measured in survey research are demographic variables, which are used to depict the characteristics of the people surveyed in the sample. Demographic variables include such measures as ethnicity, socioeconomic status, race, and age. Surveys often assess the preferences and attitudes of individuals, and many employ self-report scales to measure people's opinions and judgements about different items presented on a scale. Self-report scales are also used to examine the disparities among people on scale items. These self-report scales, which are usually presented in questionnaire form, are one of the most used instruments in psychology, and thus it is important that the measures be constructed carefully, while also being reliable and valid.

Reliability and validity of self-report measures

Reliable measures of self-report are defined by their consistency. Thus, a reliable self-report measure produces consistent results every time it is executed. A test's reliability can be measured a few ways. First, one can calculate a test-retest reliability. A test-retest reliability entails conducting the same questionnaire to a large sample at two different times. For the questionnaire to be considered reliable, people in the sample do not have to score identically on each test, but rather their position in the score distribution should be similar for both the test and the retest. Self-report measures will generally be more reliable when they have many items measuring a construct. Furthermore, measurements will be more reliable when the factor being measured has greater variability among the individuals in the sample that are being tested. Finally, there will be greater reliability when instructions for the completion of the questionnaire are clear and when there are limited distractions in the testing environment. Contrastingly, a questionnaire is valid if what it measures is what it had originally planned to measure. Construct validity of a measure is the degree to which it measures the theoretical construct that it was originally supposed to measure.