Surveys are powerful tools to develop a better understanding of your customers, target audience, employees, and stakeholders. With a rigorously designed and implemented survey, firms, government agencies, and non-profits can harness invaluable information to help inform their strategic, financial, and operational decision making. However, if design and implementation aren’t optimal, survey research can lead to ill-informed decision making based on unreliable and invalid information, and with serious consequences for operations, performance, and competitive advantage.
CEP Analytics’s team can assist you with rigorous survey development and data analysis at any step of the process. Our survey development team will provide you with a customized approach based on your research goals, sampling frame, intended output, and, of course, timeline. We’ll develop a solution specific to your organization, your survey research needs, and your sector to provide statistical consulting and analysis support for the entire survey research process.
Whether you need assistance with specifying the initial survey, developing the sampling plan, conducting psychometric tests for reliability, validity, and power, or analyzing the collected data—or each of these steps—we’ll develop a plan to meet your specific needs.
At CEP Analytics, we have vast experience with survey design and development, as well as survey research implementation and data analysis. Our survey development and data analysis team includes former professors and program evaluators who have performed survey research on behalf of the U.S. Census Bureau and Bureau of labor and statistics, alongside and strategists with long experience and expertise performing survey research in the US and internationally for major non-for-profit organizations and small and medium-sized enterprises.
If you’re developing and implementing a survey instrument to inform your organization’s operational, financial, or strategic decision making, it’s critical to plan for each step of the process. CEP Analytics’s survey development team can help you anticipate and avoid common mistakes that could render your survey datasets and data analysis invalid or even damaging, such as mono-source bias, unreliable measurement, invalid measurement, social response bias, multi-collinearity, and simultaneity bias.
Our survey development process begins with an initial consultation to understand the overarching goals for your instrument and the constructs you plan to measure. When designing a survey, researchers must identify not only how to ask questions, but what questions to ask. Our analysts will refine your information goals and determine the most precise way to measure that information. We work with you to operationalize your variables and constructs for future analysis, and to make choices such as whether or not to include open-ended or free-response questions, and when to use a Likert scale or Slider scale, along with other pre-coded response options.
Once your survey instrument is generated, we can discuss your sampling methodology and help you to select the most feasible and robust option for your research. Depending on your goals, timeline, access, and budget, participants can be selected randomly or stratified into relevant subgroups. We can help you implement purposive sampling to target specific individuals, convenience sampling to maximize ease of access, or probability sampling that reaches a random group of respondents.
Online Platform Assistance: Over the past decade, more and more of our clients, particularly in the corporate sector, are choosing to implement survey research using online platforms. We are very familiar with the leading online survey tools, including SurveyMonkey, Qualtrics, and Google Forms. If you plan to use any of these tools, we can help you input the survey into the tool for you so that you’re able to make optimal use of its data collection capabilities, as well as download and collate the data outputs once survey collection is complete. If you’re unsure of which survey tool to use, we can discuss your specific needs and goals, suggest which platform would best suit your study, and then manage the data collection process.
Pilot Testing: For clients designing an entirely new tool or making significant modifications to an existing instrument, we can help you design and then analyze data from a pilot study to test for preliminary reliability and validity. When your methodology and sampling plan are fully determined, our statisticians can perform a thorough pilot test to ensure that your survey is effective when it reaches your clients. To determine the significance of your pilot results, our analysts can check for various types of validity (such as discriminant, convergent, and concurrent validity) and reliability (such as split-half, Cronbach’s alpha, or test-retest reliability).