Features
Check your exposure to potential sample bias
Most surveys have a number of key questions, the answers to which will largely determine the business decisions taken on the basis of the results. It is therefore essential that the sample for the survey represents accurately those sub groups in the population with a significant influence on these key findings.
The Risk-E software introduces a new statistic RiskF (short for risk factor), which identifies these sub groups (known as predictors), and alerts users to the possibility of bias in the key results because the predictor groups have not been correctly represented in the sample. The data can then either be weighted to correct the bias, or additional interviews can be conducted to correct for any shortfall in numbers.
Extract key insights from your survey results
Insights from survey results often come from developing an understanding of the variables that are influencing or driving the key findings of the survey. For example, what factors are driving satisfaction or dissatisfaction with the service provided by your organisation; who is most likely to buy your new product or service?
Riske-E incorporates a DriveR analysis, which immediately identifies those sub groups that make a significant contribution to the key survey findings, and quantifies the size of the contribution.
Identify targets for marketing campaigns
The Index score compares the results of key questions (for example, likelihood to buy your new product) among sub-groups of your survey population, immediately identifying the segments with the most potential for marketing campaigns for your new product or service.
Export results to Excel
Each analysis takes only a few seconds to set up and run, so it is very easy to use to explore all the factors that may be influencing the results of your survey.
Each Risk-E analysis, including "top line" explanations of the influence of each sub group, can be exported into Excel, ensuring that results can quickly be transferred into reports and presentations.
Statistical Tests
Risk-e provides an automatic reliability estimate for every statistic and presents only results that are significant at the level of confidence. The programme also identifies results for which there is insufficient sample to determine significance or non-significance. All that is required from the user is to select the level of confidence required.
Advanced Features
Statistical procedure
For those that are technically minded, Riske-E is the first ever end-user survey application to make use of re-sampling for its statistical calculations. The statistical procedure employed is based on Central Limit Theorem, using re-sampling from the original data set.
Thus if there are 1,000 original respondents, Riske-E will select a sample of 1,000 records randomly from those respondents (with replacement) and calculate all the specified statistics on that sample. It will then repeat this process as many times as may be required for the chosen level of significance. In the case where a 95% level of confidence is chosen, the requirement is for 400 repetitions, yielding 400 slightly different values for each statistic.
The mean of those statistics, taken over all 400 re-samples, is the value Risk-E shows in its results, and the 95% confidence intervals associated with each statistic are those derived from the distribution of that statistic across those 400 re samples, excluding any outliers.
If the user requires a higher level of confidence than 95%, then it is simply necessary to choose a higher number of re-samples.
The benefit of the re-sampling approach is that users do not need to choose any statistical test methodology and so do not have to consider issues like population distribution, independence or small sample approximations.
Architecture of the technology
Risk-E is a windows desktop application and processes data provided in triple-S format, making it immediately accessible for analysis of data derived from modern data collection systems and other sources.
Reports may be viewed online, printed or exported to Microsoft Excel. They can be readily incorporated into other reports or presentations, including graphics, if required.
For a list of the software that converts data into Triple-S format please see the Triple-s website.
Further the developers have been developing a CSV to Triple-S conversion which is available for test, if required.