Data retention and research

      Data retention and research


        Article summary

        Sparx keeps school data for 2 years   

        • Parent data is deleted as soon as their child is no longer an active user
        • Personal data relating to students and teachers is kept for 2 years after they are no longer active users. This is so we can restore school data if a school re-subscribes with us or so we can answer such queries as Subject Access Requests. After this period all personal data is anonymised and kept for research purposes. You may request data to be deleted sooner if required

        Further information on our data retention schedule can be found in the Terms and Conditions > Section C: Data handling agreement > Duration of processing. We have developed our data retention policy in line with ICO & DFE guidelines. 

        We also keep anonymised data for research purposes

        • After our retention period, anonymised student data is kept for research purposes. This includes a student’s Sparx ID, their school & class, year of birth and their question-answer history or ‘Usage data’
        • Personal information such as name, UPN, date of birth and IP address are all deleted and destruction logs are maintained. As such, it is no longer personal data. The usage data is vital for us to provide personalised homework. The 100 million plus data points feed in behind the scenes to help us understand how difficult books and questions are, and how long they are likely to take students of different abilities.

        We do evaluation and product research

        Sparx uses an anonymised dataset on live students wherever possible to improve our products. When doing insights work internally or for external publication that may require one or more identifiable fields - any group of students that are small enough to be identifiable are discounted. That is to say, students would not be individually identifiable through aggregation. E.g. Disregarding international students when looking at completion rates based on location OR removing non-binary students from the study when looking at correlations of age and gender with completion.



        Was this article helpful?