The study investigated student reaction to the alternative mechanics survey (AMS), a modified force concept inventory, which used automatically marked free-response questions and offered limited feedback to students after their answers had been submitted. Eight participants were observed in completing the AMS, and they were interviewed to gain insight into what had been observed; the resultant data set was analyzed by thematic analysis. This revealed six key themes: "use of free-response questions supported deep learning", "interpretation of the AMS instructions affected answer length", "the idea of being marked by a computer did not affect answer structure", "participant reaction to the usability of the AMS was mostly positive", "reactions to the AMS depended upon what participants thought it was for", and "limited feedback was a useful addition to the AMS". Participants gave answers of differing length, being guided by the question wording as well as by their own preferences. It was found that participants valued being given feedback on their performance. Participants reacted positively to the free-response questions and could see potential for the use of this question type, opening up possibilities for the use of automatically marked free-response questions in concept inventories in the future.