Skip to main content
. 2022 Sep 30;9(3):e37894. doi: 10.2196/37894

Table 1.

Categories and subcategories of the theme “organization of recommendations,” quotations supporting the categories, and number of experts that made comments in each category/subcategory.

Categories Subcategories Citations (examples) Experts, n (%)
Hierarchization

N/Aa
  • There are recommendations with different levels of detail, and they are all placed at the same level; some recommendations correspond to guidelines, others are practical design indications. [Eb6, male]



  • It would be interesting to organize the recommendations based on the relationship between them. A high-level recommendation contains low level recommendations. [E8, male]



  • It makes sense to split into layers. I suggest dividing them into recommendations applicable to all and into standards, with a very high level of detail (button size, space between buttons etc). [E1, female]


4 (44)
Grouping of recommendations Design
principles
  • During our analysis, we organized the requirements into categories to assist us. [E7, female]



  • To be able to do the analysis, I had to code each recommendation to more easily identify the ones that could be grouped according to that category and thus detect repetitions and redundancies. [E6, male]



  • To make it easier, I created categories according to Nielsen's 10 usability heuristics and 5 design principles. [E3, female]



Of the 9 experts, 5 suggested categories for grouping recommendations:


  • Feedback; users/confusion, errors; human-computer dialogue; system behavior; navigation; presentation; system; users; instruction/information; user control; personalization; system; screen reader; users/cognitive load; system/devices; instructions; user/emotions; design considerations; gamification; users/sensory. [E7, female]



  • System status; feedback; task execution; navigation/ interaction; organization/structure of information; attention-orientation; information hierarchy; iconography; visual composition of information and of interaction elements; naming; navigation; accessibility; input device and interaction; attention-orientation; learnability; interface customization; complexity and density of information; typography/legibility and formatting; media controllers; color and contrast; navigation/task execution; information representation/mental models. [E6, male]



  • Feedback; visibility; multimodality; help; recognition; mental burden; control; design; real world; language; flexibility; errors; consistency; accessibility; personalization; search history; animation; efficiency; number of steps; shortcuts; hierarchy; legibility; color; cultural context; security; body; discovery; emotion; gamification; predictive; privacy. [E3, female]



Feedback; text information; user profile; layout; navigation; content; tasks; errors; consistency; input; ergonomic; emotional; security; gamification. [E2, female]


  • Visual dimension; dimension of information architecture; dimension of social presence; interaction dimension and dimension of user experience. [E1, female]

7 (78)

Generic vs specific to technology/ interaction paradigms
  • The recommendations focus on different types of interactive products (feet, audio interaction, robotics, etc). It would make sense in the end, too, to organize and separate the recommendations by product types. [E6, male]



  • I noticed that some recommendations distinguish interaction modalities (with voice, gestures, use of the feet, etc). I selected those that were general and those that were specific for these groups. [E3, female]



  • All recommendations were put in the same bag, regardless of the detail. I think we are using design principles, guidelines, and standards. Standards only make sense when applied to a specific system, and it is very difficult to classify them without having in mind the system being evaluated. Design standards are derived from design principles and guidelines but applied to specific products. What is being done here is very rich and interesting, but it can lead to a “Frankenstein product,” because the recommendations depend on several factors. I think that at this point you should work with principles, point out the guidelines and check the recommendations for the different products. [E1, female]



  • As we are talking about recommendations that cut across different types of interactive products (feet, audio interaction, robotics, etc), it would make sense in the end, too, to organize and separate the recommendations by product types. [E6, male]



  • They are related in the area of interaction, but each one talks about a different interface. [E3, female]



Of the 9 experts, 3 suggested categories for grouping recommendations:


  • Voice interaction; feet interaction; robot. [E7, female]



  • Generic; generic/user centered. [E6, male]

  • Voice; feet; real world; robot; touch; click; text; gestures. [E2, female]

3 (33)

aN/A: not applicable.

bE: expert.