Hi there,
I am wondering how the average value of the "Control" condition is calculated in the publication "A simple procedure for the assessment of acceptance of advanced transport telematics". If I use the three after measurement values provided in Table 5, I either get 0.5633 as average Satisfying Scale or, when using the sample size of Table2 in order to weigh the different studies, my result is 0.5602.
Both of these average values for the Satisfying Scale would however be on the right side of the "Control" condition in Fig. 2 (see also the attached picture showing a figure created by myself printed over the original Fig. 2 image).
Are not all “Control” conditions included in the value of the figure? Or is there a mistake with the values in Table 5?
Any feedback on this question appreciated a lot!
Best regards,
Thomas
Article A simple procedure for the assessment of acceptance of advan...
Wow Thomas,
That is from really really long ago, and now you are checking the figures and all I can say is that it made me frown too. It is obviously not the average of the figures given in the table. I think (but would have to find the old data...and I have no idea) that it may be that the averages in the figure are the output of spss and missing values in the within subjects design have led to slightly deviant values form the averages one can calculate on the basis f the table. So for example:
Before After
2 4
3 5
4 6
2 5
. 4
. 4
4 3
Avg: 3 4,42 on basis of 7 observations
Avg: 3 4,60 on basis of 5 observations.
Regards...Dick (and co-author Adriaan) :-)
Hi Dick and Adriaan,
Thank you very much for coming back to my question. Well, the reason I was trying to replicate your figure is that I wanted to know the exact values of your control average in order to compare these values with results of mine. Unfortunately I do not have any control condition for my experiment.
I think your explanation might point into the right direction. However, in the paper you wrote that you only used the after-measurement data for figure 2 (“Based on the usefulness scores and satisfying scores in Table 5, and the characteristics of the different in-vehicle systems evaluated, the after-measurement scores were averaged across individual experiments, which resulted in eight (pairs of) scores. These scores are displayed in Fig. 2 in a two-dimensional space.”
Maybe this applies only for the measurements regarding specific systems and for the calculation of the overall control condition you also used the before measurements? Doing this, I come much closer to the value in the figure ((0.1 + 0.27 + 0.31 + 0.3 + 1.08)/5 = 0.412).
May I ask another question regarding the methods of your experiments which is not explained in the paper: How did you ask for the control conditions as no system was used on those runs? What did you put in the question “What is your judgment about a system that would …”. Did you tell the subjects they have a system but the system was actually deactivated?
Kindest regards,
Thomas
Yes, the best one to answer this question is the authors of the paper. Dick one of the authors of the paper came and replied you.
In general,concerning statistical data,the average and the standard deviation are normally used to repentant random variables having Gaussian distribution.
May not all of the data coming out a questionnaire can be included in the statistics. Some votes may be excluded. because they are wrong .
In order to compare quantitatively statistics resulting from questionnaire, you have to ask the same questions on a similar voters with same number. Not only average values are included but also their standard deviations. Other wise the comparison may be quantitative. Also the number of voters must be sufficiently large to the general validity of the results.
This is my opinion on this question. Hope it will help.
- Badges
- Science topic
- Similar topics
- Correction