Fig 1.
Operationalized 3D-LAP coding scheme for the scientific practice “Developing and Using Models”.
We coded each exam item that asked students to engage with a visual model as being “partially aligned,” “mostly aligned,” or “fully aligned” to the 3D-LAP criteria [23] for “Developing and Using Models.” We coded items that did not ask students to engage with a visual model as “No alignment” to “Developing and Using Models.” The 3D-LAP criteria for selected-response and constructed-response items differed in how they describe student interactions with the item (e.g., the criteria indicates that students are asked to construct an explanation rather than select an explanation). The constructed-response criteria statements were operationalized in an identical manner to the selected-response criteria statements here.
Fig 2.
Alignment of cell/molecular exam items to the scientific practice “Developing and Using Models”.
Exam items are color coded to distinguish between selected-response and constructed-response format.
Fig 3.
Relationship between self-reported frequency of teaching using models and the percent of exam points that contain a visual model.
Teaching practices were self-reported on item #39 from the Measurement Instrument of Scientific Teaching (MIST) [8]. Instructors who reported that they taught using models zero times during a semester were binned into “Zero times”. Instructors who reported teaching using models less than or equal to 2-3 times per month were binned into “Low frequency.” Instructors who reported teaching using models greater than or equal to 1-2 times per week were binned into “High frequency.” Individual instructors are represented with the smaller purple dots. Larger black dots represent the mean and lines represent the 95% confidence interval.
Fig 4.
Proportion of cognitive skills assessed by items that did not and did contain visual models.
Proportions calculated based on the total number of items with no model (n = 2252) and the total number of items that contained a model (n = 435).
Table 1.
Linear regression model predicting the Bloom’s Taxonomy level of items based on alignment to “Developing and Using Models”.
Fig 5.
Box-and-whisker plots of the Bloom’s Taxonomy levels of items at each degree of alignment to the scientific practice “Developing and Using Models”.
Solid bars within each box represent the median value, boxes represent the interquartile range, and whiskers represent 1.5 times the interquartile range. Bloom’s Taxonomy levels 1–6 correspond to the cognitive skills “remember,” “understand,” “apply,” “analyze,” “evaluate,” and “create,” respectively.
Fig 6.
Categorization of the representations of DNA in molecular biology exams based on their alignment to the scale and abstraction axes of the DNA Landscape.
There were unique DNA representations on 87 exam items, and 12 of those items had representations that were coded in multiple places on the DNA Landscape.