Home Original Proposal Assignments Group Members

Cognitive Walkthrough

In order to improve the cocktail menu interface and to ensure that the actions of the interface matched user expectations, we conducted a cognitive walkthrough. We followed a cognitive review process compiled by Gregory Abowd which can be found here. We outlined tasks and corresponding actions, and proceeded to test the actions using the four following critique questions outlined by Abowd:
  1. Will the users be trying to produce whatever effect the action has?
  2. Will users be able to notice that the correct action is available?
  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
  4. After the action is taken, will users understand the feedback they get?

Summary

The cognitive walkthrough brought up a few repeat issues that must be evaluated in the high-fidelity prototype. In Actions 1 and 8, we realized we must use images that appear as buttons for the user. If not, the user may not realize specific actions are available. In addition, there were two main interaction we need to design as we move forward. First, we need to determine what happens when a user loads a previously selected drink. Will the drink's attributes be listed as the selected attributes, or will all of the attributes listed when the drink was selected re-load? Second, we need to design how adding an attribute from the drink descriptions changes the rest of the interface. Will the attribute gray out everywhere it appears, or will the attribute disappear?

The cognitive walkthrough also exposed areas that needed to be further defined. For instance, while completing the analysis on Action 2, we determined that clicking anywhere on an attribute would either expand a listing, or do nothing if the attribute could not be expanded. Attributes can only be added by selecting the green add circle on the attribute.

Cognitive Walkthrough

Task: Navigate to cocktail menu from home screen
        Action 1: Click "Cocktails" button
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes, they want a cocktail, so they will be trying to produce a list of cocktails.

  2. Will users be able to notice that the correct action is available?
    3. Maybe: We need to make sure it is obvious that the drink types are buttons, and not just a list.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Yes, this is analogous to a paper menu.

  4. After the action is taken, will users understand the feedback they get?
    5. Yes, the screens change completely.

Task: The user decides instead of popular drinks, they want to find a drink with a specific liquor.
        Action 2: Click liquor button to expand liquor attributes.
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes.

  2. Will users be able to notice that the correct action is available?
    3. Yes, assuming they see the arrow.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Maybe. There might be some question for categories that user can add (ie. Fruit/Spice).

  4. After the action is taken, will users understand the feedback they get?
    5. Yes. The category will expand, so the feedback will be clear.

        Action 3: Click add circle on rum attribute to add rum.
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes, given the initial instructions on the bottom of the attribute side.

  2. Will users be able to notice that the correct action is available?
    3. Yes. With initial instructions.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Yes, again, with initial instructions.

  4. After the action is taken, will users understand the feedback they get?
    5. No, we should design a flow of changes to describe the result of the action of clicking the add button.

Task: The right side displays a default list of drinks containing "rum". The user browses these drinks
        Action 4: Scroll through drinks by clicking and dragging the drink.
  1. Will the users be trying to produce whatever effect the action has?
    2. Questionable: Assuming we have a multitouch screen, it would. On a computer it may be difficult.

  2. Will users be able to notice that the correct action is available?
    3. Maybe: Users might not realize the list is scrollable. We need to give them a visual clue that this is possible.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Yes. Once they realize they can interact with the drink list, it will be obvious what it does.

  4. After the action is taken, will users understand the feedback they get?
    5. Yes. The drinks will move, and the drink description will change. There will also be some animation.

Task: Try to find drinks similar to the one listed.
        Action 5: Press green circle on currently displayed drink attribute to add attribute to list.
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes, they want to add the attribute.

  2. Will users be able to notice that the correct action is available?
    3. Maybe. The add circle with the plus sign is standard on our interface, but they must first notice it on the "description" listing.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Yes, if they understand the left half menu, it's the same action.

  4. After the action is taken, will users understand the feedback they get?
    5. Maybe.. We need to determine how the attributes will change appearance on the description listing for each drink, and on the attributes listing on the left half of the menu.

Task: User wants to continue browsing, but wants to remember current drink.
        Action 6: Press "Bookmark"
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes.

  2. Will users be able to notice that the correct action is available?
    3. Yes

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Maybe. Depends on what they want to bookmark. Current drink or drink and attributes list?

  4. After the action is taken, will users understand the feedback they get?
    5. . Yes. There's an explanatory popup.

Task: User wants to go back to a previous drinks
        Action 7: Press "Bookmarks" tab on right corner of the screen.
  1. Will the users be trying to produce whatever effect the action has?
    2. Yes, this is related to what they want to see.

  2. Will users be able to notice that the correct action is available?
    3. Yes, the Bookmark tab is always available.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Yes, no other actions "Bookmarks" could be tied to.

  4. After the action is taken, will users understand the feedback they get?
    5. . Yes. Bookmarked drink list appears.

        Action 8: Click on desired drink in "Bookmarks" list.
  1. Will the users be trying to produce whatever effect the action has?
    2. Questionable. We need to determine what happens to the rest of the interface (will previous selected attributes pop up? will the specific drink pop up only?)

  2. Will users be able to notice that the correct action is available?
    3. Questionable. As long as drink listings actually look like buttons.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Questionable, we need to determine whether this restores the drink or the search, and what attributes would be populated.

  4. After the action is taken, will users understand the feedback they get?
    5. Questionable, depends on how we define the action of the drink button on the bookmarked list.

Task: User wants to place an order
        Action 9: Press "Call Bartender" button
  1. Will the users be trying to produce whatever effect the action has?
    2. Questionable, user wants result of the effect (bartender comes over) maybe not the actual effect (light turns on under menu)

  2. Will users be able to notice that the correct action is available?
    3. Yes, it's completely separate from the rest of the menu.

  3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
    4. Questionable, depends on if they trust the system, but they'll know it's related to what they want.

  4. After the action is taken, will users understand the feedback they get?
    5. Yes, the light appears right after pressing the button.