Surveys Research

Surveys Research
Eric Audras/PhotoAlto/Getty Images

Surveys are non-experimental methods for collecting data on phenomena that cannot readily be observed. Surveys may employ interviews, open-ended questions, or forced-choice questions.

The two major categories of surveys are cross-sectional and longitudinal. Cross-sectional surveys are a snap-shot in time. Longitudinal surveys collect data across periods of time and are designed to capture changes or trends.

Longitudinal surveys that use the same sample of respondents are either cohort or panel studies.

Surveys Research Question Design

Both online and offline surveys can benefit from a variety of question designs. Here are several from which to choose.

  • Discrete Choice Versus Conjoint
  • Simple Branching
  • Piping
  • Looping
  • Randomization
  • Compound Branching
  • Delayed Branching
  • Simple Quotas
  • Nested or Complex Quotas
  • Extraction
  • Show/Hide Question
  • Show/Hide Answer
  • Self-Determined

Surveys Research Sampling

Surveys are based on the idea that careful selection of a sample can provide valid and reliable information about the larger target population. There are several important factors that must be taken into consideration when constructing surveys research samples.To be useful, a surveys research sample must be representative.

It must look - taste - behave like the target population. Sample data must, as closely as possible - reflect that of the target universe. Once sample data has been collected and analyzed, it is described in terms of the confidence that researchers have in the validity and reliability of the data.

Confidence in Surveys Research Sample Data

Analysis of Qualitative Survey Data

Strong data analysis "begins with the end in mind." Steven Covey may not have been the first to promote that adage, but it is often attributed to him. The strength of this approach in survey conceptualization is that one can begin with a data dashboard in mind, thereby avoiding time-consuming attempts to get the data to conform to the display later on. Treatment and analysis of survey responses differ enormously for different types of data, yet most survey data can be displayed in a tabbed-dashboard that presents information categorically and visually.

Analysis of Quantitative Survey Data

Computer assisted data analysis speeds the way to deep and useful information. Probabilistic Structural Equation Modeling (PSEM)is a faster version of SEM that uses Bayesian networks to accomplish data analysis.

  • Product Launch Survey with Product Optimization
  • Bayesian Networks for Analyzing Survey Data

Managing the Mechanics of Surveys Research

One of the most important and challenging aspects of surveys research is ensuring that the maximum number of completed questionnaires are received. A number of strategies can help keep a survey top-of-mind in a survey participant's busy day.

Reporting Surveys Research Outcomes

A good rule of thumb when designing a report format for surveys research outcomes is to let the audience determine the complexity of the content. The purpose for which the surveys research data is generated should drive the report design. The higher the stakes, the more detail and statistical evidence should be provided.

Scoring and Reporting Survey Outcomes

Types of Market Surveys Research

Surveys can be open-ended or close-ended. Survey structure or framework can be designed for online presentation, web-based data collection, or for the management of enterprise feedback. They can be directed customers, employees, or the general marketplace. Surveys are typically designed to gather information that is explanatory, predictive, or sometimes simply descriptive.

A survey may be designed to combine any or all of these three research objectives. There are many different survey methods and techniques for analyzing survey data. Several are listed in the sub-section immediately following:

Surveys Research Data Analysis Techniques

The data analysis techniques listed below all have scientific, mathematical names. But software applications exist to make the process of using the techniques quite easy and expedient.

  • Bayesian Belief Networks
  • Structural Equation Modeling
  • Probabilistic Structural Equation Modeling
  • Key Driver Analysis
  • Cross-Tabs Analysis
  • Conjoint Profile Analysis
  • Delphi Method
  • Customer Satisfaction Surveys
  • Net Promoter Scores