MarkBook® section 8-9
Class Modal Analysis for "Most-Consistent" Performance MarkBook® section 8-9
Class Modal Analysis for "Most-Consistent" Performance 
Statistically, "mode" is the most common number is a set of numbers. Applied to student marks, mode is the most frequent or consistent number in a student's set of
assessment data. There is an argument that says that a student's overall grade should be assigned, not calculated, on the basis of their "most-consistent" performance i.e. their mode. In case of a tie between two modes, the higher mode (best mode) will be selected. However, when students are
graded on a percentage scale and there are a limited number of assessments, strict modal assignment can be misleading or
blatantly incorrect. For instance, suppose a student had earned the following fifteen percentage marks: The initial impression: this is a capable student who usually performs well. The unweighted average of these fifteen marks is 79%. However, a strict interpretation of "grade assignment by mode" would generate an overall grade of 34% because 34 is the most frequent number in this set of data. The way around this problem (lots of possible marks and relatively little data) is to create "levels" each of which cover a defined range of percentage marks. For instance:
less than 50% = level R (i.e. remedial)
51% to 59% = level 1
60% to 69% = level 2
70% to 79% = level 3
80% to 100%+ = level 4
The use of Levels has an additional advantage.
They fit very well with the use of rubrics and exemplars. When a teacher grades
subjectively, it's often difficult to give a rationale for a difference of a few
percentage points (what's the difference between 81 and 82 on an essay?). Also,
two different teachers covering the same course may exhibit substantial
differences in grading standards. However, if grading is done by level with
descriptions of student performance at each level, and exemplars of that work at
each level, then the grading process is
more objective and it's consistent from one teacher to the next. Our sample student above has thirteen of fifteen marks in the level 4 range.
Once categorized into levels, this student's mode is level 4. The "most-consistent" grade assignment argument says that this student has
earned an overall grade of "level 4" despite the two marks in the remedial range. Note that the calculated average, 79%, is a level 3. Thus, in this modal method for determining an overall grade, this
student will be assigned a grade
(letter, level, or percentage) which, in this case
ignores the infrequent poor performances, happens to be higher than the
arithmetic mean (average), and best reflects this student's demonstrated capabilities. Put another way, this student's "most consistent performance" is at level 4. Therefore, "level 4"
or an equivalent percentage like 85% or 90% is reported as the overall grade because that level best reflects the student's demonstrated ability. 79% or level 3 should not be used because this student typically and frequently performs above a level 3 standard. Furthermore, this student has
no assessments in level 3 and has never performed at that level. Using MarkBook, one can determine the mode for all
assessments in the whole course (i.e. "macro").
It's also possible to use modal assessment on sub sections of a course, assign a
modal grade for that subsection (i.e. "micro"), and then enter that
modal grade into MarkBook as a single entry. Then, the overall grade will be calculated with more familiar and traditional weighting procedures. The following example uses a printed MarkBook class list grid. Print these forms as described in manual
section 5-1. Note that this teacher elected to print 20 columns with a delineating dark line every 5. This Grade/Form 9 class - 1D1 - is taking "Healthy Active Living" (HAL) - a Physical Education course. The curriculum unit is "Interactive Activities" which this school district has decided will count for 10% of the overall term mark for HAL. This teacher has hand-written four categories at the top of the form with five columns each. The decision as to how many columns to use for each category was hers. She has created assessment instruments in each category, graded
these using a rubric, and recorded individuals'
scores on each item by level, 1- to 4+. A blank cell indicates a missed item (NoMark). So far she has recorded:KnowUnd = Knowledge and Understanding 5 assessments ThinkInq = Thinking and Inquiry 3 assessments Comm'n = Communication 4 assessments Appl'n = Connections/Applications 3 assessments
Once the unit is finished, she can determine a summary grade for each of the four categories. For instance, the first student has five assessments for the K&U category.
Frank
has a tie between 4 and 4- for his most-consistent level in this
category. In case of a tie, the better level is used. The fourth student, Daisy, gets a 2-. In the circumstance where there is little consistency, e.g. Daisy in the T&I category, select the middle level 3. When an
assessment has been missed, such as for Anne Oy in the Communications category, the modality principle generates a level 4 even though an average of these four entries with one zero would likely generate a 3. Once a 1-4 level is assigned to each category, put four new entries into MarkBook. This teacher set up her five curriculum units as if they were MarkBook categories (i.e. her first MarkBook "category" in HAL was UNIT 1, her second MarkBook "category" was UNIT 2, and so on). Thus, she has four MarkBook entries per unit - one for each curriculum category. In this way she can easily adjust the relative weighting within each unit as per the following table. The exam
or "Summative Assessment" counts as if it were a sixth "category".| HAL UNIT | % of Term | K & U | T & I | Comm | Conn/
App'n |
| 1. Interactive Activities | 10% | 10% | 10% | 30% | 50% |
| 2. Physical Fitness | 20% | 20% | 20% | 20% | 40% |
| 3. Healthy Living | 25% | 30% | 20% | 20% | 30% |
| 4. Large & Small group Activities | 30% | 30% | 10% | 10% | 50% |
| 5. Body Management Activities | 15% | 25% | 15% | 20% | 40% |
| TOTAL TERM: | 100% |
| Term: | 70% |
| Summative Assessments (exam plus others): | 30% |
| TOTAL for HAL | 100% |
Unfortunately, strict modal analysis makes an assumption that is frequently erroneous: all
assessments have equal value or weight. MarkBook calculates class modal performance in three ways: the Entry
Weight Method, where all
assessments count with weight determined by the instructor and category weights
are ignored, the Equal
Weight Method, where all assessments are weighted equally, and
a Category Weight Method in which all
assessments are weighted independently of their base and category weights take
priority. For each of these, view overall performance (i.e. ALL entries) or some subset of entries
using the Unit: and Category: toggles near the top of the
following screens. In the following sample screens, the teacher has defined thirteen levels, 1- to 4+ including a remedial range (R) for anything under 50%.
Any change to this setup is done in section 7-1.
The Calculation Method is determined in the Mark Set screen (click Mark Set,
select Edit) as per the following. By default, MarkBook assumes Average and
Category Weight.
If the user has
selected Entry Weight in the prior screen, the following screen will
display. In this calculation, all assessments are
weighted by their base mark. That is, a test out of 50 and weighted
the same counts for
twice as much as one out of 25 regardless of the assessment category. Note that
the calculation method shows in blue near the top right of the next screen. The trigger or threshold value for each level is indicated in brackets underneath
the level. That is, the lowest score in each level is indicated. An entry must be equal to or greater than the threshold score in order to fit in the range. Thus, a mark of 52% would count in level 1- since it is greater than the
threshold of 50, but below the required threshold for the next level of 53. Similarly, any mark in the 90's will count in level 4+. A horizontal dotted line indicates no
assessment data in that level on this class summary screen.
Using this display, the teacher can determine which level has the modal performance i.e. the most entries. For instance,
student #15, Hap, has 54% of his assessments in the 4+ level whereas the 22nd student,
Pepper, has 41% of her assessments in the 1- level. Whether 4+ and 1-
are appropriate overall grades for these two students is a professional
judgment on the teacher's part. However, it would be prudent to check the other two modal displays as per the following
screens before making a final
judgment. The second of these modal views is an
Equal Weight calculation. Again, MarkBook presents the percentage of each individual's
assessments in each level. However, in this display, all
assessments are valued or weighted equally regardless of category. That is, a small quiz and
an examination contribute equally. This method is used in circumstances where
the kind and value of assessments has little variety.
Note that
numerous students have a different mode in this calculation compared to the
previous one above. Apparently, the class had a number of lightly weighted
assessments on which they did very well. Since this method displays mode with all
assessments equal, many students have a much higher overall level. The third view is
MarkBook's default. Experienced
evaluators know that
a unit test or major project should contribute more to the overall grade than a homework check, a quick mini test, or a uniform inspection.
As an example, suppose our student at the beginning of this section had those two marks of 34% on the mid-term and final exams
in a senior high school course. The other thirteen items were on small items such as homework checks. With this new information in mind, would a 4+ level still be appropriate as an overall grade? Obviously not! MarkBook users have a weight assigned to each
assessment entry and each category as in section 7-4. These weights are editable as in section 7-6. This third view of modal performance, Weighted
Average with Categories, factors in the weight or relative importance of each
assessment AND the weights of the categories. The largest number in each row indicates the most consistent level with weighting
and categories taken into account. As with the previous screen, MarkBook
determines each student's overall mode in the
Mode column. Note how the inclusion of
weighting alters the
Mode column for several students. To edit the weights of categories and individual entries, click on the Weight
button at the lower right. See section 7-6 for more information on the concept of weighting and how to manipulate weights.
To print any modal analysis, click on the Print tab near the top right of this screen. On the subsequent Print Class Report screen, select
Modal Analysis at the bottom. See section 8-6 for setup details. To see an individual's modal performance
as a bar graph, go to section 9-4. To exit from this print screen, click on the Back button. Or, go to the
Weight screen or the Anecdotal Notes screen using those buttons.Asylum Software's Home Page | Send Us Email | LockerManager | MarkBook CNX for handheld computers | Why Use Class Management Software? | Start of the MarkBook Manual | Order a printed MarkBook Manual | LogiNotes (MarkBook En Français) | Bar Coded Attendance | MarkBook's Home Page | Electronic Report Cards | Order MarkBook