| Function |
Graph choice: participant is explicitly stating graph choice (i.e., bar, line, scatter) based on the data provided in the table. Participants may also interject their personal feelings or rely on their past experiences when contemplating between different graph types, their usage, and limitations. |
GS1: Oh that’s a good point, whether or not I can connect them, because [with] the [variable] time line can be discrete. I’m not sure. I think since its cell growth over time that should be fine [to do] so (connects points on the graph). UGR2: I’m using the line graph because it shows the trend the easiest, because it goes straight and up a little. UGR3: … did I say line or bar? I’m doing lines. I’m doing a line chart now I changed my mind. |
| Invention |
Statistics: participant is talking about either descriptive or inferential statistics. |
P5: [the trend] is almost linear and [this is] because there is some error [in the data] which I didn’t calculate (sketches error bars on each data point). P2: You do need a bigger sample size, but [I will estimate] the error [bar] for each one [treatment]. (adds error bars and labels lines as either 10C or 22C). GS7: I think what I’m going to do is take average of three tubes and make a bar for each time point at each temperature. I’m plotting to show the standard deviation from the average value. UGR4: … you can create a trendline for each dataset, so basically out of 15 ml and 5 ml, you can do the line of best fit, where you try and roughly go through as many of the points as possible. UGNR7: This graph looks like it’s not going to be linear, but I’ll make a line of best fit for each [tube] just so you can tell where it’s going. |
|
Data type: participant is explicitly making decisions about whether or not to plot raw data or plot manipulated data (i.e., average) and the number of graphs to use to properly convey the data. |
P1: I’m collapsing across tubes, so I’m giving total [number of cells], or I could do mean [number of cells]. GS5: There are three tubes within each temperature group, so I will do the average—calculate the mean of the number of cells for the same time point for all three tubes. And for each time point I can have the mean and standard deviation. UGR3: Okay well I’m going to make two charts then if that’s the case. I’ll make one the cell count at 22 degrees Celsius, and I’ll make another one for cell count at 10 degrees Celsius with the same axes. UGR4: Because we have three plants, which is like three trials for each, I’m going to average the number of leaves at each time for each plant for each amount of water. UGNR6: I’m thinking maybe I could do like an average number of plants that would require doing calculations. There’s fifteen milliliters of water a day. I’m just going to go ahead and do averages. |
| Learning/reflection |
Evaluation: participant is talking either about the general graphing habits, future directions, or take-home message. |
P2: You do need a bigger sample size, but [I will estimate] the error [bar] for each one [treatment] (adds error bars and labels lines as either 10C or 22C). GS8: This is the most horrible graph ever because it’s not even clear what the data mean. It might be easy for me to understand what I’ve done but it’s not easy. If I gave it to you, I’m sure you would not understand it, if it was out of context. UGR4: You can see really clearly that they [lines] are increasing at the same rate but throughout the entire experiment, the 5 ml produces less leaves. UGNR3: I did this wrong … I should have put ml on the y axis … I’ll just keep going with this. I might be okay … okay yeah I need to plot this with number of leaves instead of ml [scratches the x-axis label and renames it number of leaves]. The number of leaves will be on the x axis. |
|
Technology: participant is mentioning the habitual use of graph-making software to reflect on elements of the current graph construction. |
GS3: So if I read the problem and use Excel, I can just put linear regression lines and the r2 values, both are greater than 0.8 or something (draws 2 linear regression lines through points and labels lines with r2 > 0.8). UGR1: So I feel like if I was doing this in Excel, I would make each plant its own representation symbol or its own color to better represent that. Have like a uniform structure to this but a different representation. UGR4: … if you are in Excel, you can [get] the equation for the trend line and it will tell you that y equals some function of x. From that, you can see the mathematical relationship behind the number of leaves that you have. |
| Critique |
Aesthetics: participant is using elements of graph design (i.e., gestalt principles and color) to critique the constructed graph. |
UGR2: I guess I will graph the other [tubes] too and we can just imagine that they are different colors. UGNR1: I’d use different colors for the ones at 22 [degrees Celsius] and the ones at 10 [degrees Celsius] and then you can show that in the legend … . But the legend is black so I guess I’ll just graph the points at different lines. They will all be the same color. |
|
Sample size: participant is critiquing the small sample size presented in the data table. |
P4: With 3 plants in each, I guess you could put a standard error on that [data point]; n = 3 is pretty small but sometimes in biology, you are stuck with pretty small. I can’t [calculate standard error] in my head but, what I would probably do is put each standard error at each [data] point, plus or minus. P2: You do need a bigger sample size. UGNR6: I’ll draw the dotted line that represents the five milliliters of water per day, which is also approximately a linear line but if there was more data it could possibly be curving off to give a constant average, at least if you want any of those. |
