R.A.L. Consulting Limited provides a full range of survey research services.  We work with a network of experienced associates including Dr. Joseph Fletcher of the University of Toronto and Gerald Bierling of McMaster University. 

Our surveys offer detailed insight into the perceptions of your target population.  We work closely with you to find out exactly what you need to learn from your respondents.  We can offer economical mail-out surveys as well as state of the art computer-assisted telephone and Internet based survey services.  Whatever method is chosen, the same level of custom service is provided.

Custom Design in Survey Research

Every aspect of the survey work we do at R.A.L. is tailored to fit your needs.  Starting with the design of a study, we work closely with your project team to understand your concerns.  We take the time to listen, ask questions, and learn what you really need to know from survey research.  This same spirit carries throughout all phases of a study -- from the collection of data to the analysis and reporting of results.  In the data collection phase, for example, we can do more than simply ask a series of questions.  We can probe your respondents' needs in ways that reveal their subtlety and nuance, not just the broad outlines.  In our reports we do more than summarize a set of percentages; we interpret the overall pattern of findings, placing the results in their proper context through the use of easy to understand graphics and charts. 

Our focus is upon answering your planning questions scientifically, using statistically sound techniques and methods of analysis.  In clear, concise terms our report will help you understand the dynamics behind your clients' perceptions by drawing concrete linkages between specific responses and their underlying values and demographics.  Our product, in short, is a fully customized approach to survey research, specifically designed to meet your needs and answer your questions.

Questionnaire Design

The design of your questionnaire is perhaps the most crucial part of a survey.  Great care must be taken to ensure that the questions ask what you really want to know.  We will work in close collaboration with you to develop a set of instruments that will yield useful data to guide your deliberations and assist in planning decisions.  Content will be will be worked out in collaboration with your project team, keeping an eye to both current concerns and possible comparisons with other available data.

Sample Design

The characteristics of the target population and budget shape our approach to sample design.  The simplest approach is send a questionnaire to every respondent in the target population.  While this has the obvious benefit of offering every respondent an opportunity to take part in the survey, this approach is also the most expensive.   A less expensive approach is a simple random sample but this approach can produce poor results in areas or groups with low response rates or few potential respondents.

A more efficient and usually our preferred sampling design is to combine the best features of the two approaches in what is called stratified sampling.  The central idea is to distinguish between those groups, or strata, that have relatively few members and those who have substantially more. Compared with surveying everyone, the stratified approach essentially reduces the overall sample size while maintaining a viable number of respondents in the smaller areas or groups.  Compared with the simple random sample, this approach concentrates on the number of respondents in the areas where they are most needed for the effective analysis of the data.  The potential savings on mailing and data entry costs using a stratified sample can be substantial.  These savings are, of course, achieved at the expense of some loss in accuracy of the estimates for areas where respondents are most numerous, but the accuracy remains high in these cases, exceeding that of any for the smaller areas.  Appropriate weighting of the data will permit estimates accurate at approximately plus or minus 3%.

Data Analysis

In analyzing the data from the survey we will use SPSS, the most widely used statistical analysis package.   With it we will perform the necessary frequency distributions, demographic breakdowns, as well as prepare other relevant cross-tabulations.  In addition, we have broad experience with advanced statistical techniques and their effective presentation, including Multiple Regression, Logistic Regression, Analysis of Variance, Factor, and Cluster Analysis.  These techniques can aid in understanding the underlying structure of the survey responses. 

Since tables and other numerical displays are often hard to digest, we set out the findings in easy to read charts and graphs created through PowerPoint.  Our consultations ensure that the final report will feature the types and depth of analysis you require and presented in the graphic formats you prefer.

Presentations and Report

Included in our survey services are two presentations and a written report.  The first presentation is of preliminary results.  After the survey data have been analyzed, but before we write the final report, we will make an informal presentation to the project team and other interested parties.  This presentation offers a first look at the results and an opportunity to raise questions and issues for further analysis.  The second presentation is more formal and can be made available to a broader audience as it is based upon the results and interpretations contained in the final report.

The report is a full written account of the survey, its results and their interpretation, complete with charts and graphics.  Going beyond the simple tabulations usually offered in reports, we provide a thorough understanding of the data and what they mean for your concerns.  This understanding is based on solid statistical analysis, not quick generalizations from a frequency distribution or cross-tabulation. 

Along with the report we make also deliver the full data file, set up as either an SPSS system file and or rectangular file in ASCII format according to your preference.

Return to previous page

© R.A.L. Consulting Limited, 2008