Talk:Applied Data Management Charts and Graphs

Question about mean versus average in representing data

I am working on assignment _________

I believe it is best to represent the data as a horizontal bar graph. Each bar should give the viewer an impression of the way the various communications were rated by users. I think an average of the scores for each would make more sense than the mean of the scores. One abnormal high or low score could skew a mean whereas an average is more reflective of all those surveyed.

To demonstrate that those surveyed are only a sample group out of a larger group, it is understood that there is some variances from how one sample group might respond to the next. The standard variance represented as a 5% deviation, or 95% confidence level is going to be used. So for any given bar in the graph, showing the average score, should also have a portion at the end shaded, or in another color to indicate that deviation which is calculated derived from the confidence interval upper, and lower value.

For example, for Video tapes & DVD the bar should indicate the average score with the end portion of the bar graphic indicating that variance on confidence. I believe I will take the computed average, and subtract the lower confidence level to indicate the beginning of the alternative shade on the bar, and the upper level to indicate the end of the shaded portion of the bar. So the result will be a bar slightly longer than it would be just using the average.

What I am uncertain about at this point is the use of average versus mean. It seems that the confidence interval is calculated using the mean while the average should be used to indicate the original bar length prior to the confidence interval overlay. Is this correct?

• Mean or average for initial graph representation of results?
• Confidence interval overlay on each bar utilizing computation based on mean and confidence level?