Helpful Authoring Definitions
In the Create an Item tutorial, we created a multiple choice Question. In this tutorial, we are going to take a closer look at some of the scoring options covered in that tutorial.
Most of Learnosity’s Question Types are autoscored - excluding Essay, Audio and Image Highlight questions. Each auto scored Question Type has validation settings. Authors can select the scoring type, enter the points for the Question, enter a penalty score for the Question, and enable a Check Answer button.
Scoring Types for Questions within an Item
In most cases we can validate questions using 3 scoring methods: Exact Match, Partial Match Per Response, or Partial Match. Some questions, such as Classification and Order List, have additional scoring methods. More details on these can be found in tutorials for these individual Question Types.
The Exact Match scoring option is available in all autoscored Question Types. With Exact Match, the student must get the entire Question correct to receive a mark.
The image above illustrates a Match List question scored using Exact Match. If each response is answered correctly, the student will receive a score of 1.
Partial Match per Response
Partial scoring can be used in Questions where the student may have to answer a Question with multiple responses. Each response can be scored independently, however each response is of equal value. These scoring options are not available in all of our Question Types.
If the Question is given a score of 1, as shown below, then each response within this Question will be worth 1 mark. As a result, a possible max score for the Question will be 3.
Partial Match is similar to Partial Match per Response. A student gets a partial score for each correct response within the Question. The difference with Partial Match is that the value entered in the Points box is the overall mark for the Question.
If the Question is given a score of 1, as shown below, then each response within this Question will be worth .3333 (rounded to 4 decimal places).
Under Partial Match there is also a Rounding option. Rounding is only available when the scoring type Partial Match has been selected. There are two options to select from the drop down menu: None or Floor. None is set as the default option.
- None will round the score to four decimal places e.g. 2.6666
- Floor will round the score down to the nearest whole number, e.g. a score of 2.666 will be rounded down to 2.
This will determine if marks are deducted for an incorrect answer. By default, the Penalty Score is set to 0. In the Match List examples above, there are 3 responses to the Question. If we apply a penalty score of 1 to the examples shown above, then:
- Exact Match - If the student answers any of the responses incorrectly, they will receive a score of -1 for the Question.
- Partial Match Per Response - For each incorrect response, the student will receive a score of -1, with a possible max penalty score of -3.
- Partial Match - For each incorrect response, the student will recieve a score of -.3333 with a max penalty score of -1. If rounding has been enabled, it will also apply to the penalty score.
Check Answer is a button that appears below a Question, and allows students to check if their response is correct. It is usually only applied to formative assessments. Check Answer can be enabled by toggling the Show Check Answer Button option in the Question Editor. Feedback attempts can be configured to allow the student to test as many times as needed. The default setting is unlimited attempts. The Check Answer button will not disclose the correct answer, but will indicate which answers have been answered incorrectly as shown below.
Non Autoscored Questions
Essay (including Math and Chemistry), Audio and Image Highlight Question Types need to be scored by a teacher. While a correct answer cannot be specified in the Question Editor, a maximum score can be set. The option can be found under the More Options accordion.
The Unscored/Practice Usage option allows authors to disable scoring for a question. When enabled, the question will receive no score. This can be found in the scoring section of a question.
Math Specific Scoring
- The equivSymbolic method checks that the value entered by the student is mathematically equivalent to the value that the Author has set in validation, even if they are in different forms.
- Use equivSymbolic when working with equations, or other input with variables, where order or form is not important. For example, if the task is to multiply 3 and 2, any response that is mathematically equal to 6 will be considered as correct, such as 6/1, 12,1, 3+3, 6.0
- equivSymbolic accepts decimals, fractions, variables, and percentage.
- The equivLiteral method checks if the response from the student is literally equal to the correct value specified by the Author. This means that the form, order, elements, and values entered by the student should match the value in validation.
- Responses that are mathematically equal, but are in a form other than the one specified by Authors, will return false.
- Use equivLiteral when you want to strictly specify a particular form of the mathematical expression as the correct answer.
Note that equivLiteral ignores parentheses which have no specific meaning, by default.
- The equivValue method compares numerical values that may be represented in different ways, such as units of measurement, where 1 m = 100 cm. equivValue will evaluate the expression to a numerical form for comparison. As long as the value given by the student is mathematically equal to the result the author was expecting, the question will validate as correct.
- Although similar to equivSymbolic, the equivValue method cannot be used with variables. It does however work with decimals, fractions and percentage. equivValue is particularly useful if you are using units of measurement. This method also handles validation of lists and ranges, using the tolerance symbol.
Scoring types for the Item as a Whole
Learnosity supports three different types of Item scoring: Per Question, Dichotomous, and Dependent (EBSR: Evidence-Based Selected Response). The scoring type you select determines how the Item will be scored in a user's session.
This is the default scoring type. The score for the Item is calculated by adding the score from each Question on the Item.
For the purposes of the following demos, a fully correct Question scores 10, a partially correct Question scores 5, and an incorrect Question scores 0.
Figure 1: Per Question Item scoring.
Dichotomously scored Items have only two possible outcomes: fully correct or fully incorrect. If the user's score for every Question on the Item is fully correct, the score for the Item is calculated by adding the max scores from each Question. If any Question on the Item is not fully correct, the score is zero.
Figure 2: Dichotomous Item scoring.
In dependent scoring, the score taken from each Question depends on the scores of previous Questions on the Item. The rule is that if a Question on the Item is not fully correct, the user is not awarded a score for any subsequent Questions on the Item.
Such scoring can be applied when there are two or more Questions dependent on each other. For instance, the first Question has to be answered correctly because the following one is based on how the student got to the right solution.
This scoring type can be used to emulate an Evidence-Based Selected Response (EBSR) Item type.
Figure 3: Dependent Item Scoring
To adjust this scoring method:
While you are in the item...
Step 1: Click on Settings
Step 2: Select Details
Step 3: Make any desired changes. The scoring options can be selected from the Scoring Type dropdown menu. Click HERE for more information on scoring options for items in an advanced assessment
We covered common scoring components across all question types, however, the specific method of adding validation can vary between Question Types. Click here for more information on individual question type scoring.