Help Needed

You are correct when you state that there appears to be a linear relationship between the variables. In your posts you also stated that, "It is correct to say longer viewing times result in greater preference for ratings as compared to shorter times." This phrasing is not quite accurate. By stating the the times result in greater preference we are indicating a causal relationship - which is not something we can determine from this study. Instead, we need to limit our interpretation of the results based on our basic understanding and state that we can see there is a correlation between the amount of viewing time and aesthetic rating scores. If we wanted to be able to make a causal statement about the variables in this study, how could we develop a research design that would allow for that?

The scatter plot suggests a positive relationship between viewing time and aesthetic preference. On average, as one increases, so does the other. The amount of scatter shows a fairly strong correlation, but not perfect. Perfect correlations usually do not happen. The correlation could be calculated numerically by the correlation coefficient.

 Are longer viewing times the result of greater preference? We cannot make that assumption without a true experiment that involves temporal precedence and controls for third variables. We cannot prove that one factor is the result of another with the information we have now. We are only showing a positive and fairly strong correlation not causation. It could be the opposite, where greater preference is the result of longer viewing times, or there can be many confounding third variables not being identified such as crowding around certain paintings, or seating arrangements.

*The scatterplot range shown above is a positive correlation with entire range of values shown in the increase of data with viewing time and preference rating.

Scatter plot graphic representation of each individual’s scores on two variables. The score on the first variable is found on the horizontal axis and score on the second variable is found on vertical axis. For example the scatterplot relationship in this graph seen between viewing time and aesthetic preference shows the correlations between both the rating and viewing time. The x axis shows the rating and the y axis shows the viewing of the time. In the data shown it does show a positive increase in the value of the viewing time and preference rating, however it is not a perfect diagonal showing the variation between the two relationships. The values of the first variable are depicted on the x axis, and the values of the second variable are shown on the y axis, which does not show as a perfect positive relationships between both variables. I am still trying to analyze the data presented to accurate relate to how efficient the positive scoring is viewed.

Class, how can you tell the difference between a positive correlation and a perfect positive relationship? Is there a difference between the two?

Y-Values 10 12 24 5 16 3 11 5 21 23 9 3 17 14 3 4 7 3 6 4 4 2 8 9 5 3 5 6

Page 2