Having created a prototype, we then polled a selection of potential users for the interactive fridge using this questionnaire. We assigned two tasks for the users to do, these were to find a recipe for pancakes from the computers internal database and to find out what produce the fridge was running low on. The results of these tasks will be used to help evaluate the interface design of the system. In addition to these two tasks, we asked the users to evaluate the functionality and design of the system.
Results:
1) “On a scale of 1 to 10, where 1 is “Not clear at all” and 10 is “Very Clear”, how clear is the main menu page?”
All of the test users found that the main page was clear & easy to understand. Some of the suggestions were:
- As this is a touch-screen system, the buttons should be as large and as clear as possible to make use as easy as possible.
- It would be beneficial to put a clock on the main menu.
- The main page is under-used, it would be helpful for the system to give additional information on/next to the buttons, for example “a shepherds pie recipe suggestion” next to the recipe button.
All of the test users found that the main page was clear & easy to understand. Some of the suggestions were:
100% of the results rated the page as 8 out of 10 or above on clearness.
The mean result was 8.86, median 9, mode 8 (with a sample of 7 people).
The standard deviation was 0.83
From these results, we can conclude the main menu page has been well designed, however there is some further functionality that several users would have liked to have seen implemented.
All of the test users found that the main page was clear & easy to understand. Some of the suggestions were:100% of the results rated the page as 8 out of 10 or above on clearness.The mean result was 8.86, median 9, mode 8 (with a sample of 7 people).The standard deviation was 0.83From these results, we can conclude the main menu page has been well designed, however there is some further functionality that several users would have liked to have seen implemented.
2) “On a scale of 1 to 10, where 1 is “Very difficult” and 10 is “Very easy”, how easy was it to find a recipe for pancakes?”
Every user found this task extremely easy to achieve without prior knowledge of the system. It took a maximum of three clicks to locate the correct page.Every user found this task extremely easy to achieve without prior knowledge of the system. It took a maximum of three clicks to locate the correct page.
100% of the users rated the task as 10 our of 10 for ease of use.
The mean, median and modal results were 10 (with a sample of 7 people).
The standard deviation was 0
From these results, we can conclude the process used to find recopies was well designed.
Every user found this task extremely easy to achieve without prior knowledge of the system. It took a maximum of three clicks to locate the correct page.100% of the users rated the task as 10 our of 10 for ease of use.The mean, median and modal results were 10 (with a sample of 7 people).The standard deviation was 0From these results, we can conclude the process used to find recopies was well designed.
3) “On a scale of 1 to 10, where 1 is “Very difficult” and 10 is “Very easy”, how easy was it to find out what items you were running low on?”
There was a greater spread of results for this task, most of the users found the information in the end; several could not find it at all.
The mean was 5.14, the median was 6, and the mode was 1 (with a sample of 7 people).
The standard deviation was 2.90
A number of users believed the required page would be located in a different category, either the ‘inventory’ or ‘status’ pages. From these results, we can conclude that this information should be available in both places. According to the users, it should be located in the inventory as it is the most intuitive places to look. It should also be available on the shopping page so users can order what they are running low on.
4) The rest of the questions required qualitative answers, here are the general conclusions we have made from the results (available on the questionnaire results page):
- As the design is to be implemented as a touch-screen system, the on-screen buttons should be as large as possible. The data-input interface should be made as easy as possible to use.
- The need for the status page in its present format was often questioned as it ‘didn’t seem to really do anything’. To one user, its existence seemed only to provide a gimmick with the thermal imaging camera. (however this may provide a useful marketing strategy). To another user, it should have been easier to understand more explicit information regarding the temperature of individual compartments, rather than a colored image. One user suggested an overview of the information displayed on this page could be used as a ‘screen saver’ giving important information (such as the day, date & time, as well as current temperatures, meal suggestions and warning notices), rather than as a specific page the user has to manually navigate to. Fahrenheit and Celsius should be available as units for displaying the temperature to the user.
- Several users found the recipe feature pointless as they either already have a range of recipes at their disposal in books, or were experimental cooks and don’t like using them. Another user would only want it if it only gave you suggestions based on the ingredients available to them from in the fridge. It would be helpful if these ingredients could be purchased in a job-lot if required. Users would like the ability to download more recipes from the internet and arrange them using their own categories.
“What might be nifty, […], is the ability to get it to read the recipe to you, and set timers for stuff on it (eg, sauté the onions for x minutes. Press the button when you have started.” “The onions are sautéed. Take them off the heat now.”) - The most popular feature was to sort via use-by date, as users liked to know what items should be used up first. This ties in with our original questionnaires results indicating people wish to reduce the amount of food they waste.
- Several users raised the concern that manually removing items from the fridge’s database as they are used would require an excessive amount of work on a day to day level to be feasible. They would only use this feature if an automated system were implemented to keep track of food usage for them.
This problem was not addressed in the prototype as the problem is related to the mechanical functionality of the fridge, rather than the management software designed. Previous ideas suggested for the fridge dealt with this problem by using pressure sensors and RFID tags to keep track of what was used. RFID’s were discounted as a number of people raised concerns over the potential invasion of privacy, abuse and other security problems. - A large number of users expressed the desire of remote access to the fridges database. This would be beneficial for users at the supermarket, or to help them plan what they are going to eat that evening. This functionality could be provided via the mobile phone network or an internet connection using an authentication system.
A more accurate system for tracking exact quantities of the amount of produce stored within the fridge would be beneficial (e.g., the fridge may say you still have butter (enough for 1 sandwich, but not enough to bake a cake).