Merge branch 'MDL-42743' of git://github.com/timhunt/moodle

question/behaviour/deferredcbm/behaviourtype.php

@@ -40,6 +40,7 @@ class qbehaviour_deferredcbm_type extends qbehaviour_deferredfeedback_type {
     }
 
     public function summarise_usage(question_usage_by_activity $quba, question_display_options $options) {
+        global $OUTPUT;
         $summarydata = parent::summarise_usage($quba, $options);
 
         if ($options->marks < question_display_options::MARK_AND_MAX) {
@@ -107,6 +108,13 @@ class qbehaviour_deferredcbm_type extends qbehaviour_deferredfeedback_type {
         $cbmbonus         = $this->calculate_bonus($averagecbm, $accuracy);
         $accuracyandbonus = $accuracy + $cbmbonus;
 
+        // Add a note to explain the max mark.
+        $summarydata['qbehaviour_cbm_grade_explanation'] = array(
+            'title' => '',
+            'content' => html_writer::tag('i', get_string('cbmgradeexplanation', 'qbehaviour_deferredcbm')) .
+                    $OUTPUT->help_icon('cbmgrades', 'qbehaviour_deferredcbm'),
+        );
+
         // Now we can start generating some of the summary: overall values.
         $summarydata['qbehaviour_cbm_entire_quiz_heading'] = array(
             'title' => '',

question/behaviour/deferredcbm/lang/en/qbehaviour_deferredcbm.php

@@ -31,6 +31,12 @@ $string['basemark'] = 'Base mark {$a}';
 $string['breakdownbycertainty'] = 'Break-down by certainty';
 $string['cbmbonus'] = 'CBM bonus';
 $string['cbmmark'] = 'CBM mark {$a}';
+$string['cbmgradeexplanation'] = 'For CBM, the grade above is shown relative to the maximum for all correct at C=1.';
+$string['cbmgrades'] = 'CBM grades';
+$string['cbmgrades_help'] = 'With Certainty Based Marking (CBM) getting every question correct with C=1 (low certainty) gives a grade of 100%. Grades may be as high as 300% if every question is correct with C=3 (high certainty). Misconceptions (confident wrong responses) lower grades much more than wrong responses that are acknowledged to be uncertain. This may even lead to negative overall grades.
+
+**Accuracy** is the % correct ignoring certainty but weighted for the maximum mark of each question. Successfully distinguishing more and less reliable responses gives a better grade than selecting the same certainty for each question. This is reflected in the **CBM Bonus**. **Accuracy** + **CBM Bonus** is a better measure of knowledge than **Accuracy**. Misconceptions can lead to a negative bonus, a warning to look carefully at what is and is not known.';
+$string['cbmgrades_link'] = 'qbehaviour/deferredcbm/certaintygrade';
 $string['certainty'] = 'Certainty';
 $string['certainty_help'] = 'Certainty-based marking requires you to indicate how reliable you think your answer is. The available levels are:
 
@@ -52,8 +58,8 @@ $string['certaintyshort1'] = 'C=1';
 $string['certaintyshort2'] = 'C=2';
 $string['certaintyshort3'] = 'C=3';
 $string['dontknow'] = 'No idea';
-$string['foransweredquestions'] = 'For just the {$a} answered questions';
-$string['forentirequiz'] = 'For the entire quiz ({$a} questions)';
+$string['foransweredquestions'] = 'Results for just the {$a} answered questions';
+$string['forentirequiz'] = 'Results for the whole quiz ({$a} questions)';
 $string['judgementok'] = 'OK';
 $string['judgementsummary'] = 'Responses: {$a->responses}. Accuracy: {$a->fraction}. (Optimal range {$a->idealrangelow} to {$a->idealrangehigh}). You were {$a->judgement} using this certainty level.';
 $string['howcertainareyou'] = 'Certainty{$a->help}: {$a->choices}';