SA823(JAN11) : ACADEMIC DISCOURSE
For Topic 6. Writing the results section
Organisation of moves and graphic representation of evidence in Findings
This wiki records observations on how the Findings or Results section of published research papers is organised in different disciplines, what moves are made, how they the moves are typically arranged, and how graphics for communicating findings are presented. Please state the discipline/subject area in which the observations were made
Discipline / Subject Area
Legitmizing Computer Simulations to support Context for Doing / Physics Education
Title of Article: When learning about the real world is better done virtually: a study of substituting computer simulations for laboratory equipment
Authors: N.D. Finkelstein, W. K. Adams, C. J. Keller, P. B. Kohl, K. K Perkins, N. S. Podolefsky, S. Reid, R. LeMaster
Methodology: 3 steps-rich triangulation research method, giving a real sense of the learning situation while maintaining research vigor
experimental group CCK N=99 perform lab with computer sim,
control group TRAD N=132 perform lab with real equipment,
null group No Lab N=107
Draft 2 for writing discourse.
Organization of Findings
Draft 2 for writing discourse.
The findings (results) are presented in two sections, A. Circuit Challenge and B. Themes from observations. But only A. Circuit Challenge has graphics thus will be discuss here.
Use A. Circuit Challenge and arrange findings into 3 parts 1 Timing: a performance task on connecting up a circuit suitable, 2 Student writup: an established and accepted means of assessing whether students are able to provide logical and scientific explanations of the circuit and 3 Final exams: a pen and paper assessment held 2 months after the intervention lesson of 1.5 hours to demonstrate the learning in not a short term effect. A timing performance of a real life task, pen and paper cognitive assessment, interviews with participants allows for the study to be interpreted as scholarly.
Organization of Findings (III Results)
A. Circuit Challenge
A detail and honest context setting account of the study is presented in a typical university courses illustrating the complexity and authenticity (as close to real conditions in learning in university courses) setting of the study.
In 1 Timing, the moves are presented in a graphical bar chart arranged from reading logical left to right format, the experimental (CCK), control (TRAD) and null group (No Lab), with y-axis representing the time in minutes taken to complete the constructing of a real circuit.
Findings are communicated by displaying consistent colors of the graphical bars with typical associated error bars illustrate very visual representation of data. In the paragraphs, the W use statistical method to suggest experimental group CCK perform task in shorter time than TRAD while excercising care not to over claim about the scholarly data presented. A subsequent follow-up argument is used by means of a statistical technique and assumption to argue greater statistical significance difference.
W ends paragraph move by drawing on supporting findings of shorter time to construct the sample circuit challenge task in both CCK and TRAD, and compare with No Lab group to illustrate the relatedness of CCK and TRAD as suitable means for context of learning by doing. This is a very sound move given that the physicist education community has high regard for learning with real equipments like TRAD.
2 Student writeup
In 2 Student writeup, the moves are presented in a graphical bar chart arranged from reading logical left to right format, the experimental (CCK), control (TRAD) and null group (No Lab), with y-axis representing the percentage of population sample size. Again, W gives rich account of the context of the student writeup task to demonstrate research procedure in a honest fashion.
Findings used in paragraphs are in averages of the rubric scores differs from the graphical rubrics scores in percentages, gives an alternative statistical favor to the findings perhaps to anticipate R's scholarly lens. W argues of a statistics difference in flavor of CCK in statistical language and representation. W continues to advance his/her claim that the findings suggest CCK students were better able to integrate concepts, a necessary ability for the performance of this writeup task. Though some R may not be particularly convince of the W's claim in this particular task 2 Student writeup, it is clear that most R can appreciate W's scholarly moves made.
3 Final exams
In 3 Final exams, the moves are presented in consistent graphical bar chart arranged from reading logical left to right format, the experimental (CCK), control (TRAD) and null group (No Lab), with y-axis representing the fraction of population sample size.
Findings are selectively done with choice of questions in final examination from inferring that the research task is strongly correlated to q1,q2 and q3. W also anticipate R possible questions by also including the final examinations score labelled as "cntl".
W moves to arrange in statistically representation of the findings to advance W consistent argumentative claim, as consistently thorough all 3 parts. W did anticipate the final exam score as not statistically different between CCK and TRAD but later argue that it is within scholarly expectations not have difference. By doing this, W uphold research integrity while still advancing W paper's claim.
An excellent paper!
Organization of Findings III Results
A Circuit Challenge
Set the context for task of the challenge circuit to build a circuit using real equipment in student working groups of two to five students, as shown in FIG 2. Schematic diagram on the main challenge task as the common task for all 3 groups. Show to Teaching Assistant TA then timing is recorded and the circuit is broken at the switch where they will answer a short essay question on the circuit and turn in their answers individually.
The graph is clear y-axis as time taken in minutes to conduct the challenge task (build, broke the circuit and answer questions), the bar charts had 3 groups CCK, TRAD and no Lab, with the associated 95% error bars. At one glance, the data clearly flavors the CCK in completion of task in shortest time.
Use statistics method to illustrate control group CCK perform task in shorter time than TRAD but being careful about the scholarly data presented, but also use a technique and assumption to argue greater significance.
CCK average time = 14.0 minutes
TRAD average time = 17.7 minutes
No Lab average time = 26.7 minutes
Argue that CCK - TRAD < 0 significant p < 0.1 level ( two tailed t-test )
use a technique of including data fastest and slowest group within each section ans assume these times serve an upper bound estimate, find
CCK - TRAD < 0 significant p < 0.01 level ( two tailed t-test, pooled variance )
hence, CCK - No Lab <0 and TRAD - No lab <0 is even more significant.
2 Student writeup
Set the context for task on writeup after the circuit is broken, what to describe, why bulbs change brightness, can use formula or words, use everyday language to perform a ranking task, which is brightness etc..
A standardized rubrics with scale 0 (no demonstrated knowledge) to 3( correct and complete reasoning) was used
research team reach consensus for inter-rater correlation, for correctness and use of concepts, mathematics, illustrate thoroughness of scholarly study.
Simple fraction was used to report the figure 4
average score CCK = 1.86
average score TRAD = 1.64 argue of significant shift p<0.03 two tail z test.
also qualify what is not reported here (use of mathematics)
Use significant difference between CCK and TRAD to suggest CCK students were better able to integrate concepts, a necessary ability for the performance of this writeup task.
3 Final exams
Use back the same Fig 2 of earlier task.
Set the context for final exam performance task.
Spilit into 3 questions
q1 rank the current through each bulbs
q2 rank voltage drop across bulbs
q3 predict through bulb 1 increased, decrease remain same brightness when switch is open.
cntl remaining 26 question of exam cover other materials of the course
fraction correct on cntl
CCK = 0.621 (N=99; σ= 0.18)
TRAD = 0.612 (N=132; σ= 0.17) statistically identical
mean on q1 q2 and q3
CCK = 0.593 (N=99; σ= 0.27)
TRAD = 0.476 (N=132; σ= 0.27) argue of significant difference on circuit q1,q2 and q3 (p<0.002)
the way the data is present is still consistent with the main rhetorical goal.
An excellent paper!