Evaluating the Effects of Intervention on Teacher Learning in an Online Heat and Temperature Course.
I. Analysis of online discussion forums: Radiation Session (session 5)-Cohort One
Part A
1. Did the teacher make a heating prediction and was the heating prediction (for the lamp heating the cans) consistent with scientific norms? (Black can heats up at a faster rate and heats to a higher temperature than the white can. The white can heats up but at a slower rate and to a lower temperature.)
All twenty-five teachers posted their prediction about the heating of the black and white cans. Twenty-four teachers predicted the black can would heat faster and to a higher temperature than the white can. One teacher wrote,
"Black Can-Prediction temperature will increase at a decreasing rate curve on a graph. White Can-Prediction temperature will increase at more constant rate on a graph. The amount of heat the black and white shirts would absorb is different white gathers/absorbs heat more slowly that black."
From this observation, its difficult to assess what this teacher thinks will happen. Does she think the white can will get as hot as the black can, just more slowly? Does she think the heating of the white can is linear ("more constant")?
2. Did the teacher make a cooling prediction (after the cans had been heated by the lamp) and was the cooling prediction consistent with scientific norms? (The hotter black can will cool faster than the cooler white can.)
Twenty-one teachers made predictions about which can would cool faster. Seven teachers predicted that since the black can would get hotter than the white can, and it would cool faster. Fourteen teachers predicted that the black can would heat up faster and stay hotter longer. Many teachers stated this claim was due to their experience with black car interiors or black clothes.
3. Did the teacher make a cooling prediction of the cans starting at the same temperature and was the cooling prediction consistent with scientific norms? (The black can will cool faster than the white can. Both cans decrease in temperature at a decreasing rate with the black decreasing faster.)
Fourteen teachers predicted that when both cans started at the same temperature of around 45 degrees Celsius, that the black can would cool at a slower rate than the white can. Three teachers reported that they thought the black can would cool faster then the white can. One teacher referred to the Hewitt text in explaining his thinking about why the black can will cool faster, "A good absorber makes a good emitter." Eight teachers did not make a prediction about how the cans would cool when cooling from the same temperature.
4. Did the teacher report results for each test and are the reported results for each test consistent with the scientific norm?
In the heating-by-lamp experiment, nineteen teachers reported that their results matched their predictions and their black can heated at a faster rate and to a higher temperature than the white can. Seventeen of these nineteen teachers also reported results that were consistent with the "ideal" results; the black can cools faster than the white can. Seven other teachers reported that their findings were inconsistent with their predictions and either there was no real difference in heating and cooling.
5. Is there at least one Logger Lite graph posted to the report?
The majority of teachers posted at least one of the graphs as instructed, either the heating graph or the cooling graph. Five teachers did not post any graph, and it was clear from the reporting of two teachers, that the experiment was not done.
6. Is the graph labeled and consistent with the claims reported in the written report?
All the graphs except one were unlabeled. Many people did not include any indicator in their report to explain what color line (red or blue) referred to which color can. It is unclear whether the teachers knew which can corresponded which graph. Although this was the third science session in which the Logger Lite software was used, these teachers had not determined how to annotate or label their graphs or data. What was most surprising was that there was only one teacher who queries her group, "I want to know how to label the legend on the graphs so that you can know which is which just by looking, not having to keep going over to the table. Do you know?" No one responded to her plea for help, not even the facilitator.
Part B
Problems occasionally arose with the design of the investigation and the equipment used in session. There were missed opportunities for deep learning due to the lack of presence in the forum by the facilitators, great variability in participants' ability to stimulate vigorous discussion among colleagues, a reluctance of many teachers to challenge unsupported assertions, inadequate presentation of data, and an inadequate use of the graphing software. It was also discovered that the investigation designed for the session on radiation was problematic due to the use of painted cans full of air. The cans had a screw-on lid with a hole in the top for the temperature probe. Teachers reported that their cans fell over, the lids may not have been on with the same tightness, and in one graph, it was clear that the probe in the black can had tipped over and was taking the temperature of the aluminum instead of the air inside the can. Interestingly, when this graph was posted, no teacher noted the sudden and rapid decline in temperature.
When data did not match predictions, the teachers often tried to salvage their predictions; at other times they interpreted their data as being wrong (apparently because they did not agree with the prediction):
I am conflicted about posting this week's results because I am really confounded by them and I keep thinking they are incorrect.
I worked on this with a colleague who is feeling the same way. We ran the experiments at school. We both predicted that the black can would heat up faster and stay hotter longer because of our experience with black clothes in the summer. They feel warmer than white clothes, right?! (Though what about the black, full-body covering clothes worn in the desert?! Is the reason it's worn because of the type of fabric or the color?)
To our dismay, in both experiments it was the white can that got hotter, though the black can cooled off more quickly in the lamp experiment. I wondered what accounted for this, and frankly, I am still wondering. I feel like one of the young children in my class who is perfectly willing to change the results of an experiment that they ran in order to meet their own prediction so that the event make sense to them. (This is one of the dangers of science with young children that makes it so fun, and interesting to work with them on. It promotes the need for good teachers of hands-on science, yet I find myself behaving like a first grader with this experience and so far no one can guide me through my muddled thoughts. I can't wait to see some of the other results in our group!) I was thoroughly surprised to see that the white can heated up more than the black. This was totally counter-intuitive to me. I wondered if there was some mistake, so I ran the experiment again. And I was careful to triple-check the cables to make sure I was really looking at the curve for the right can.
However, I am wondering if some variable threw off our work. I noticed the white can's lid was tighter than that of the black can. Maybe some of the air was able to escape out of the black can because of the loose lid? Also, when I did the experiment the first time, a little blob of clay melted down (unnoticed by us at the time) into the bottom of the black can. Did this cause conduction of heat to the bottom of the can making it cooler than the white can?
I wonder if the molecules that are being heated at the top are staying at the top, and affecting only some of the molecules near the top of the can. Could there be less hot molecules at the bottom of the can, which are not heating up?
Further reading finds that these two teachers used a "very hot utility lamp" instead of the 60 watt bulb and lamp provided in the course two kit. The fact that the white can reaches a higher temperature indicates that it is "blacker" - i.e. a better absorber - at the visible and near-infrared wavelengths given off by the heat lamp than it is at the much longer infrared wavelengths at which it radiates in order to cool off. This was not a concept that the teachers were prepared to understand. This was not a concept intended to be taught during this session. Concepts of visible and infra-red wavelengths come in Fulcrum course three.
II. Analysis of Pre and Post Responses to a Thought Experiment (Cohort One and Cohort Two)
Preliminary data from the Cohort 1 pre and post assessments show an increase in the total number of ideas used by the teachers to explain the phenomena from initial assessment to final assessment, but curiously, there was a decrease in the number if ideas of certain microscopic concepts that we might have expected to increase. While we found a large increase in the use of the ideas (numbers 2,4, and 6): temperature differences between substances; conduction/convection; and phase change to explain which system will cool the soda the fastest, we found there is a decrease in the use of the idea (number 8): speed of molecules of the substance in the system. One hypothesis for the decrease from pre to post test in the use of phrases associated with the speed of molecules has to
do with the thought experiment itself; the teachers were instructed to, "Imagine you have a pair of magic eyeglasses that allow you to observe the system on a microscopic scale." In the pre-test teachers might have felt obligated to use these terms, even with little understanding. But without showing how the ideas are linked, it's difficult to compare the complexity of ideas in both the pre and post-test. There is an increase in the use of the idea of kinetic energy to explain heat transfer at the molecular level in the post-test. Future analysis would include trying to develop a method to show the complexity of the ideas.
Based on a preliminary random survey of seven teachers pre and post-test responses, six of the seven used more ideas to respond the post-test than in the pre-test. For four of the seven teachers, their ratio of normative to total ideas increased in the post-test. One teacher increased her number of ideas to explain her reasoning in the post-test, but had a 1.0 ratio of normative to total in both the pre and post-test. What this method leaves out is the total times an idea is recorded in one response, and the complexity or linkages that these ideas may have to each other. In the future we hope to develop a snap shot of these teachers both at the beginning and end of the course that will show that their repertoire of ideas has increased as well as their ability to think through the complexity of the problem.
Comparative research between Cohort One results and Cohort Two results will take place once Course Two is completed in January 2008.
