The evaluation type eval numeric does a numerical comparison of the student’s answer against the expression in the Definition field. There are two options to this evaluation type:

  • Range: the system checks whether the answer is within the closed interval [ range low , range high].

  • Precision: this field defines the accepted absolute value of the margin of error for the numerical answer.
    Example: If the correct answer is 1, you can use a precision of 0.1 to accept all answers in the closed interval [ 0.9 , 1.1 ].

Some examples of eval numeric used in negative feedback are given below, but we can also use this evaluation type in positive feedback or the solution definition.

The rule below accepts all answers in the interval [1.50 , 1.55]:

The rule below accepts all answers in the interval [2.51 , 2.53]:

If both range and precision are specified, the system only looks at the range specification and ignores the precision specification. If only one side of the range specification is given and the other is left empty, then the empty value is treated as zero. For example, the rule below accepts all values in the interval [ 0 , 4.5 ]:

Some additional notes on this evaluation type:

  • Variables can be used in both the range and the precision specification.

  • Before the student’s answer is compared to the expression in the definition field, both are converted to their canonical normal form using the Maxima function radcan. See this article for more information on radcan.

  • The smallest floating point number that is considered is 2^(-14).

  • If you want students to give numerical answers including a unit, you can use the evaluation type eval units.

  • This evaluation type does not have automated feedback.

More on evaluation types

An overview of all evaluation types can be found here (for mathematical answers) and here (for text-based answers). More detail on the different fields of a feedback rule can be found here.

Did this answer your question?