Table 1.

The 26 heuristics (H) present in our evaluation technique.

Category description and numbera		Heuristic
Notice or Awareness: These heuristics concern what people are told about how their data are used before their personal information is collected, such as the terms of service or privacy policies.
	H1	Before data are shared with a remote actor, the entity collecting the data is explicitly identified.
	H2	Before data are shared with a remote actor, the uses of the data are explicitly identified.
	H3	Before data are shared with a remote actor, the potential recipients are explicitly identified.
	H4	The nature and means of the data collected are explicitly identified.
	H5	Steps taken to ensure confidentiality, integrity, and quality of data are explained.
	H6	For those of above satisfied, notice is sufficiently explicit.
	H7	Can control when data are used for nonoperational secondary use, such as marketing or research.
Choice or Consent: These include the controls people have over use of their data, such as whether to permit secondary uses of their data, including marketing.
	H8	Consent acquired before data shared with remote actor.
	H9	Consent is explicitly opt-in: no preticked checkboxes, etc.
	H10	Can choose which data types are automatically collected from sensors or other sources, for example, connect a finance app to a single bank account or track steps but not heart rate.
	H11	Data collection consent is dynamic: if new types of data are being collected, consent is renewed in situ.
	H12	Data processing consent is dynamic: if the purpose of processing changes, consent is renewed.
	H13	Data distribution consent is dynamic: if the actors’ data are distributed to changes, consent is renewed.
	H14	Consent to store and process data can be revoked at any time: with the service and any other actors.
	H15	Can control where data are stored.
Access or Participation: These address issues such as whether people are able to view the data they have provided and whether they can verify its accuracy in a timely manner.
	H16	All raw collected data can be extracted from the service (in-app or via vendor’s website).
	H17	All data are available in standard text formats (CSVb, XML, JSONc, GPXd, etc).
	H18	Data extraction is available from within the service, for example, without raising a request with support.
	H19	Programmatic access to data is possible, for example, app programming interfaces are exposed.
Social Disclosure Usability: These relate to the usability of interface elements that allow users to share data with third-party services, for example, social networking sites.
	H20	Privacy controls are per-disclosure, for example, individual workouts can be published to a social networking site, not relying solely on global defaults.
	H21	Privacy controls allow granular sharing of data types, for example, when sharing a workout, the distance can be shared but not the pace.
	H22	Error prevention: is explicit confirmation acquired before a disclosure?
	H23	Minimize user memory load: Effects of a disclosure are visible throughout the disclosure flow (ie, memory of earlier decisions not required).
	H24	Minimalist: During the disclosure flow no extraneous information (such as adverts or irrelevant user interface elements) is displayed.
	H25	Consistency: Information shown during the disclosure flow is consistent with the effect of the disclosure.
	H26	Help and documentation: Contextual help with making privacy decisions is available.

^aSee Multimedia Appendix 1 for the scoring criteria.

^bCSV: comma-separated values.

^cJSON: JavaScript Object Notation.

^dGPX: GPS eXchange Format.