Analyze data as you go It is likely that at some point during this course, your group has collected your data and begun your analysis only to realize that you had misread a direction, made a measurement error, or forgot to measure something important. You might have had to backtrack or possibly even start your measurement over. This could have been avoided if you had analyzed your data as you went along. (Unfortunately, for formatting purposes the lab manual always presents the Analysis section after the Measurement section. Our measurements aren't usually too involved, so we don't lose too much, but in more complicated experiments this Measurement Corner is more crucial.) Analyzing your data before taking additional data is a good habit. One advantage of taking data in media res is that you can catch errors quickly. For example, suppose you are doing an experiment to measure how the terminal voltage of a battery depends on the resistance of one resistor in a parallel circuit. You have predicted that a larger resistance should increase the terminal voltage. You connect a DMM in the circuit, but mistakenly have the DMM in current mode instead of voltage mode. You take a few data points while increasing the resistance and you notice that the reading on the DMM is decreasing. If you are analyzing your data as you go, checking that it makes sense, it will be apparent that something is wrong. Once you have caught the error, you can correct it without losing too much time taking faulty data. Another advantage comes about because your data analysis can tell you how much data you need to take. In Measurement Corner #11, you saw how you can calculate a running average of your data. Once the running average stabilizes, no more data is needed. Likewise, if you are using your data to make a graph, once it is clear that the graph is linear, or whatever shape you expect, you have enough data. You can stop taking data at this time, but you wouldn't know to stop if you hadn't analyzed the data.

Temperature (' C) Laver thickness (Angstroms) 87.75 39.677 87.14 39.675 86.62 39.673 86.11 39.667 85.60 39.659 85.08 39.600 84.56 39.537 Figure 1: Data from liquid crystal layer thickness experiment