EPHMRA Webinar: Clean Raw Data and Reliable Crosstabs - How To?
This event has passed
Meeting Registration - Zoom
Speaker: Evgeny Zlatkovsky, N'counter
Convenor: Alexander Rummel, Aurum Research – LDC Committee member
In market research, great value is placed on data quality. Insights are the basis for making decisions about important questions for products and services. Good and correctly applied data cleaning is necessary for valid, meaningful results. The aim of data cleaning is to improve the data quality and thus the validity and informative value of survey results. This webinar will focus in the first part onThe various parameters which can spoil the data (e.g., doublets, “speeders”, outliers, improper open-ended entries) How to find them in the raw data and How to exploit them for a sustained improvement of the results.
However, a clean data set is only a precondition for data quality. Between collecting data in a raw data set and their interpretation in a report, there is an obligatory stage of data evaluation and analysis. Whereas the advanced statistics are perfectly suitable for a profound insight into each question and each probable correlation, the usual analysis goes with a bivariate analysis which aims to display all the questions at once in a simple and well-arranged way, provided with percentages and all other basic parameters, and considering/comparing several subgroups. Mostly, a cross-tabulation is the only source for a data insight, without any approaches using alternative methods. Thus, it is even more important to be confident about the values within a cross-tabulation.
This webinar will focus on the various criteria which are crucial for the correctness of the data in a cross-tabulation, such as bases, scales, routings, missing values, cross-breaks.
There will be three aspects of checking for each of the criteria above:How to check if they are correct or not How to find data errors How to clean them adequately.
There will be a live demonstration of those processes of checking, error-finding, and cleaning, using the examples of real data and cross-tabulations.
The webinar is open to any member who is interested in data as well as data professionals.