diff --git a/addons/survey/models/survey_question.py b/addons/survey/models/survey_question.py
index 01f621907fa..a976772cabf 100644
--- a/addons/survey/models/survey_question.py
+++ b/addons/survey/models/survey_question.py
@@ -345,14 +345,14 @@ class SurveyQuestion(models.Model):
                     or question.triggering_answer_id is None:
                 question.triggering_answer_id = False
 
-    @api.depends('question_type', 'scoring_type', 'answer_date', 'answer_datetime', 'answer_numerical_box')
+    @api.depends('question_type', 'scoring_type', 'answer_date', 'answer_datetime', 'answer_numerical_box', 'suggested_answer_ids.is_correct')
     def _compute_is_scored_question(self):
         """ Computes whether a question "is scored" or not. Handles following cases:
           - inconsistent Boolean=None edge case that breaks tests => False
           - survey is not scored => False
           - 'date'/'datetime'/'numerical_box' question types w/correct answer => True
             (implied without user having to activate, except for numerical whose correct value is 0.0)
-          - 'simple_choice / multiple_choice': set to True even if logic is a bit different (coming from answers)
+          - 'simple_choice / multiple_choice': set to True if any of suggested answers are marked as correct
           - question_type isn't scoreable (note: choice questions scoring logic handled separately) => False
         """
         for question in self:
@@ -365,7 +365,7 @@ class SurveyQuestion(models.Model):
             elif question.question_type == 'numerical_box' and question.answer_numerical_box:
                 question.is_scored_question = True
             elif question.question_type in ['simple_choice', 'multiple_choice']:
-                question.is_scored_question = True
+                question.is_scored_question = any(question.suggested_answer_ids.mapped('is_correct'))
             else:
                 question.is_scored_question = False
 
diff --git a/addons/survey/tests/test_certification_flow.py b/addons/survey/tests/test_certification_flow.py
index 65205a5a531..2c97fc36dbd 100644
--- a/addons/survey/tests/test_certification_flow.py
+++ b/addons/survey/tests/test_certification_flow.py
@@ -115,6 +115,25 @@ class TestCertificationFlow(common.TestSurveyCommon, HttpCase):
         self.assertEqual(user_inputs.scoring_percentage, 87.5)
         self.assertTrue(user_inputs.scoring_success)
 
+        # assert statistics
+        statistics = user_inputs._prepare_statistics()[user_inputs]
+        total_statistics = statistics['totals']
+        self.assertEqual(
+            sorted(
+                total_statistics,
+                key=lambda item: item['text']
+            ),
+            sorted(
+                [
+                    {'text': 'Correct', 'count': 2},
+                    {'text': 'Partially', 'count': 1},
+                    {'text': 'Incorrect', 'count': 0},
+                    {'text': 'Unanswered', 'count': 0},
+                ],
+                key=lambda item: item['text']
+            )
+        )
+
         # Check that the certification is still successful even if scoring_success_min of certification is modified
         certification.write({'scoring_success_min': 90})
         self.assertTrue(user_inputs.scoring_success)
@@ -197,12 +216,22 @@ class TestCertificationFlow(common.TestSurveyCommon, HttpCase):
 
         statistics = user_inputs._prepare_statistics()[user_inputs]
         total_statistics = statistics['totals']
-        self.assertEqual(total_statistics, [
-            {'text': 'Correct', 'count': 1},
-            {'text': 'Partially', 'count': 0},
-            {'text': 'Incorrect', 'count': 0},
-            {'text': 'Unanswered', 'count': 0},
-        ], "With the configured randomization, there should be exactly 1 correctly answered question and none skipped.")
+        self.assertEqual(
+            sorted(
+                total_statistics,
+                key=lambda item: item['text']
+            ),
+            sorted(
+                [
+                    {'text': 'Correct', 'count': 1},
+                    {'text': 'Partially', 'count': 0},
+                    {'text': 'Incorrect', 'count': 0},
+                    {'text': 'Unanswered', 'count': 0},
+                ],
+                key=lambda item: item['text']
+            ),
+            "With the configured randomization, there should be exactly 1 correctly answered question and none skipped."
+        )
 
         section_statistics = statistics['by_section']
         self.assertEqual(section_statistics, {