In the second part of our multi-part validation study, together with Robert Sobotka, Managing Director of Telemark Marketing, we conducted a survey on the subject of cars. Thereby, the satisfaction and expectations as well as which car brands and drive concepts are currently used, were examined.
Our approach in this study was to collect the Austrian market shares of car brands as well as engine types (e.g. gasoline, diesel, electric,...) in a survey and to compare them with the real data of the registration statistics (registration data Statistics Austria). The data of more than 200 respondents were analyzed with the help of ReDem, which should lead to a reduction of the deviation and thus to an improvement of the result.
Therefore, our hypothesis was as follows: The result is closer to the reference data with the help of ReDem's calculation.
The initial situation of the study at a glance:
- Data source: Google Surveys
- Sample size: 200 (Austrian population)
- Reference data: Car registration statistics Austria (Statistics Austria)
The results:
Question 1:
Do you own a car? If yes, please tell me the make of your car?
Blue = Unweighted result (each answer has equal weight), Green = Result after ReDem, Yellow = Reference result car registration statistics Austria (Statistik Austria)
In the first example, the result was improved by a total of 21.9%, which, as in the previous part of the validation study on banks, represents again a clear reduction of the error in the result. This improvement is composed of the average of the percentage improvement of each answer option.
Question 2:
Your car has a ...
Blue = Unweighted result (each answer has equal weight), Green = Result after ReDem, Yellow = Reference result car registration statistics Austria (Statistik Austria)
In our second question, the error reduction was even higher. By 36,7% the result was improved by the use of ReDem.
This validation study impressively demonstrates that our technology enables a significant improvement in results and brings them closer to reality.
Thus, as in the first part of the validation study, we can again clearly confirm our hypothesis.
