Hi SciFi1,
Thank you for sharing your data. I think I can help you with your graph.
Title: A Comparison Between the Total Number of Questions Answered Correctly and the Average Total Timings of All the Age Categories
This is a nice title. One minor point, though: since you only have two age categories now, you are comparing "Both," not "All."
X-axis:Tests ("Noisy" and "Quiet")
Y-axis:Number of Questions Answered Correctly (out of 33) and Timings (minutes)
This is the best way to present your data. The main independent variable (test condition) is your x axis, and the dependent variables (the things you measured) are shown on the y axis. Excellent!
Adults (21 to 50 Years of Age) Test #1: 29.9 36.3
Test #2: 30.4 31.3
Young Adults (13 to 20 Years of Age) Test #1: 28 51.1
Test #2: 29.2 50.5
Now for the presentation.
I plotted these points on the graph, and connected a line through the points. For instance, wtih Microsoft Excel, I plotted 29.9 and 28 to form one line. Then, I plotted 30.4 and 29.2 to form another line. I did this to see if one line showed more of a decrease; I interpreted this as the test which showed that more mistakes were evident (which was Test #1- the "Noisy" test). On the same graph, I compared the timings by plotting the points in the same way I plotted the points to form the first two lines. I found that the Average Total Timings for Test #2 ("quiet" test) formed a steeper line, and the Average Total Timings for Test #1 ("noisy" test) formed a line that extended over a steady increase. Does this mean that Test #2 required less time to complete? Did I even do this right?
You have presented your data in an interesting way. However, line graphs are usually used for what we call
continuous variables. Continuous variables are things that change in a continuous way, usually numerically. For example, time and temperature are continuous variables. Your independent variable, however, is what we call a
categorical variable. You have two categories: noisy and quiet. There is no real "distance" between them, so the slope of your lines are artificial - they depend on how far apart Excel graphed your categories.
The better way to present your data would be to create a
grouped bar graph. Here's what I mean. You will still have "noisy" and "quiet" as your categories on the x axis. You will still graph number of questions answered correctly and test timings, but these will be shown as vertical bars (the height of which will correspond to the values). You can use colors to distinguish among your independent variables. For example, blue can be for number of questions and red can be for time. Light blue and red can be for your young adults, and dark blue and red can be for your adults. This will show, in one graph, everything that you are trying to show.

Excel should be able to create a graph like this for you, but please let me know if you have trouble, and I will help you figure out how to do this in Excel.
On the same graph, I compared the timings by plotting the points in the same way I plotted the points to form the first two lines. I found that the Average Total Timings for Test #2 ("quiet" test) formed a steeper line, and the Average Total Timings for Test #1 ("noisy" test) formed a line that extended over a steady increase. Does this mean that Test #2 required less time to complete? Did I even do this right?
What you're trying to do here is say something about whether your results are
significant. This is a special term used in statistics to describe whether scientific results could be the result of chance, or whether there is a true difference. Here is a link to a Wikipedia article explaining statistical signifiance:
http://en.wikipedia.org/wiki/Statistical_significance To be able to say whether the differences you saw were
significant, you can perform a simple statistical test called a
Student's t-test. Here is a link to a Wikipedia article explaining Student's t-test:
http://en.wikipedia.org/wiki/Student's_t-test
Excel can perform a t-test for you, if you have the Data Analysis Tool Pack installed. (You may need your Microsoft Office CD to install the tool pack, but it does come as a standard "add-in" with Office). You would perform one t-test for each comparison. According to your data, you have four comparisons between your noisy vs. quiet tests: (1) # Questions correct for adults, (2) Timings for adults, (3) # Questions correct for young adults, and (4) Timings for young adults. To perform a t-test in Excel, arrange your original data (not the averages) into columns: one column from the "noisy" test, and one column from the "quiet" test. Select two columns at a time, for each of your four comparisons. From the Excel menu, select "Tools" and then "Data Analysis..." Because you used the same people for your noisy and quiet tests, you will use a "
paired t-test." Excel will give you a table with a lot of information. The information you are most interested in is the
P value (two-tailed). If this value is less than 0.05, then your results are considered statistically
significant. If the P value is greater than 0.05, then your results are not considered statistically significant. However, you can still say that you observed a distinct
trend towards fewer questions correct and longer timings for the noisy test in both age groups. I have suggested that you look at the "two-tailed" P value, because it is more conservative. However, if you originally predicted that you would see fewer questions correct and longer timings for your noisy test, you could use the "one-tailed" P value instead, which is a bit more likely to be significant.
I hope that helps. Don't hesitate to post again if you need more help!
Best,
Heather