Survey Design & Data Analysis Workshop

Course details

A 3 day program for someone responsible for conducting or overseeing a survey research program, whether the focus is customer satisfaction, employee satisfaction, training program evaluation, customer product needs, or another area where survey research is appropriate.

Day 1

• Survey Project Management
• Definition of a scientific survey and key terminology
• Types of surveys: periodic (relationship), transaction-driven, one-time ad-hoc
• Key stages of a survey project and requirements of the project manager
• Factors that support success in a survey project
• Exercise: Developing a Statement of Purpose and how it drives the project flow
• Questionnaire Design (Instrument Design)
• The questionnaire design process: The steps to follow
• Types of questions: demographic, attitudinal, and attributes of the relationship being measured
• How to identify the questions needed on the survey to achieve the research objectives:
• Role of focus groups, interviews, and other research techniques
• Criticality of pilot testing (pre-tests)
• Critical issues in design of a valid questionnaire:
• Identifying common errors in question construction that lead to instrumentation bias, , double-barrel questions, ambiguous wording, recall expectations, loaded wording
• Elements of a questionnaire: Introductions, instructions, initiation, section headings, and closings. Formatting a questionnaire.
• Question sequencing and branching
• Controlling for biases the respondent brings to the process: response bias
• Exercises: Identifying attributes to measure on a survey and classifying them
• - Critiquing a Survey Instrument (with lots of problems)
• - Developing Scales
• - Drafting survey questions

Day 2

• Questionnaire Design (continued)
• Review of various question formats (open-ended text, multiple choice, ordinal, interval, and ratio) and their effective use
• Impact of question format on data analysis potential
• Creating interval-rating scales and anchor selection, including Likert scales (or Likert-type scales). Key considerations in scale construction.
• Special topics: Techniques for generating actionable data and measuring importance
• Administration of the Survey Instrument
• Comparison of the advantages and disadvantages of the various administration methods:
• - Telephone surveys, IVR surveys, postal mail surveys, email surveys, and online web surveys (Internet surveys)
• Discussion of administration biases and how to minimize them
• Response rate requirements for statistically valid results
• Statistical confidence given a number of responses
• Create the survey sample: Review of various sampling techniques
• Techniques to increase response rates and minimize non-response bias
• Specific issues and concerns for telephone and online survey administration methods

Day 3

• Analysis of the Survey Data
• What to do before you start analysis: Setting up the spreadsheet and data cleansing
• Options for analyzing the data
• Generating descriptive statistics using Excel
• Bivariate and multivariate statistical analysis options with specific treatment of correlation and regression analysis
• Determining whether survey scores across groups are significantly significant differences: t-tests and Chi-square tests
• Pivot tables to generate analysis
• Statistically identifying the importance of various attributes to the survey group's overall view of the organization
• The hands-on part of the Data Analysis day is focused on attendees analyzing their own data set. Bring your laptop and a data set and you will apply the techniques we learn. Don't have a data set yet? We will have a sample data set that you can analyze. As noted above, we have learned that most people are not comfortable doing this analysis in the group setting. But you are certainly welcome to do the analysis.

You will learn how

• Create a survey questionnaire design that generates more actionable & valid data
• Use all the potential survey question formats to meet your research needs
• Write good survey questions, especially for Likert interval rating scale questions
• Understand statistical confidence, survey sampling & how to get higher response rates
• Reduce the biases your current survey practices may be introducing into the survey data set, corrupting the value of the data
• Find the full message in your survey data
• Apply the proper statistical procedure to answer a survey research question
• Be more efficient in doing the survey analysis
• Determine whether a change in survey scores over time is just by chance or is a statistically significant change in operational performance.

Updated on 15 August, 2017