-
Notifications
You must be signed in to change notification settings - Fork 0
Food For Thought Documentation
This is a horizontal prototype that encompasses features that have evolved through two iterations, aiming to demonstrate basic interfaces and interactions for viewing and creating traditional recipes. (add a link here )
- Access the prototype from https://food-for-thought.000webhostapp.com/food%20for%20thought/index.html
- Open the food for thought folder and launch the template.html
- From this page enter the app by clicking the Enter button.
- Users are free to explore the recipes that are displayed on the homepage.
- Click on the icon left of the search area opens the main menu.
- Using the icon to the right of the search area allows users to return to the home page where ever they are on the app.
- Users can submit their own recipes via the submit recipe option on the main menu.
- On the recipe pages the user can read the story and recipe, add comments and view the variations.
As food is constantly being adapted to be made in newer forms of cooking, preserving tradition is often not a priority and it is not uncommon for the traditional methods of creating incredible meals is lost to time.
The application works as a library for traditional recipes, allowing users to create and edit pages for their traditional food by submitting requests. On the page of traditional recipes, they can interact with each other through comments and by sharing their personal method of making the traditional food as "variation" recipes, which aims to create a sense of community and enhance cultural identity for people eating their traditional food. The application also includes a "family group" feature that enables users to create their family recipes with their family members, where the recipes they share in the variation section are documented with a timeline.
Food for Thought has gone through an extensive design process in order to reach the point that it is at now. This process began with the research that needed to be conducted in order to answer the questions we had - why would there be a need to preserve the traditional recipes of various cultures (after all if we weren't able to answer this question, then Food for Thought was essentially pointless). Through our research we came to the conclusion that people didn't necessarily want to preserve food; it was far more apparent that people wished to simply preserve their culture and food played a considerable role in a number of these cultures. To further complement this research, we decided to have interviews with our target audience to see if this was correct. Luckily each team member had a different cultural background, which we used to our advantage to see how different cultures wished to preserve their traditional recipes.
Initial interview and results see here
From this, we decided that an app that could preserve these traditional foods (especially their origins and their original recipes) would be an ideal way of reaching our goal.
After the initial interviews, we started drawing up what this potential app could look like:
Our plan for Food for Thought was an app that could allow users to post recipes with their origins and original recipe and other users could view them and learn about them and their respective cultures. The primary target audience of Food for Thought was determined to be young adults and older who were interested in either their own culture or other cultures (with a smaller target audience being people interested in foreign cuisine). Additionally, we looked at similar apps which would be used by this target audience to see what potential features we could implement within Food for Thought.
We decided to go with a 'globe' style approach (similar to Google Earth) upon which each country would hold each of its' respective recipes. This would be accompanied with a filtering system (which would allow users to filter by dietary requirements and country) and also a search bar. Additionally there would be a menu which provided users with a number of options (access to their account, settings, etc.). On each individual recipe page, users could see either information regarding the recipe or the variations posted by other users. These concepts were formalised into the first prototype:
User testing was conducted - and the reception to the first prototype was mixed; while the concept was praised, the execution was questioned. The overall prototype had a number of issues - but the most dominant one was the fact that a globe-style approach would be very difficult to circumnavigate, especially if the number of recipes were large in a smaller country.
Another issue that was brought up was the issue of how to determine what the 'traditional' recipe was. If one user submitted a recipe but it wasn't the true traditional method, there would be no way of stopping this (other than people leaving comments pointing out the mistake).
To fix the issues of the first prototype, it was decided that it was necessary that the 'globe' style approach needed to change. As the primary issue was the way the recipes were found, we decided that a more forum-style approach (similar to Reddit) would be necessary. Furthermore, an upvoting system would also be implemented in order for the most popular traditional recipe variation to be seen as the one to be presented on the main page for the recipe itself. After taking these issues into account and applying these changes, the second prototype was developed:
The reception to this prototype was much more positive - the results of the user test showed that by presenting the recipe in this way, the recipes are much more accessible and the simple layout of the recipe screen provided a high level of accessibility as well. A further improvement was also suggested for the actual logic of the application - as there are often a number of meals which are essentially the same, but made in a different country. The suggestion was to take both of these recipes, because although they were highly similar, they both had a significant difference in terms of cultural history.
The third prototype was presented at the showcase - and while being fairly similar to the second prototype in terms of functionality, it was intended to be a horizontal prototype and covers more aspects such as the capability to submit recipes which we have not tested before. It was built from scratch in order to create a more authentic prototype, rather than one built from a prototyping service. Several key functionalities were included within this prototype, such as the capability to add recipes and to actually write proper comments.
The results from interaction with the showcase prototype yielded a number of useful improvements. It was suggested that the filtering system provide a time filter - which would fit as several of these recipes are highly dated and being able to sort via time fits thematically with this prototype. Another suggestion was to try and better encapsulate the social aspect of the application - while the concept of variations was a start, it could be improved significantly.
- Built the high fidelity prototype
- Assisted with the digital prototype on marvel
- Planned the contextual interviews for initial research
- Created digital prototypes with Marvel
- Planned the testing of prototype one and two
- Responsible for the aesthetics of the project