...
Task Given | Unexpected Responses |
|---|
1 | - Wondered if there was a correlation between the star used to favorite the posting and the stars used in the line graph.
- Thought something else should have happened besides the star next to it as a visual reminder. Specifically the fit number changing.
|
4 | - Easily navigated to the post review screen but was confused that there was no marker for internship versus normal position.
- Did not put the term internship into the textbox.
|
5 | - Used the text box as expected, but commented that he wished that he got more feedback as he typed, requesting similarity to Facebook and Google's incremental search.
- Said that he expected a search result screen rather than a direct linking to John's profile page.
|
6 | Performed as expected. |
...
Task Given | Unexpected Responses |
|---|
1 | |
2 | - Took clicked on the link of the job almost immediately, but was surprised by the slide-out behavior as the "is normally for minmal additional data, not the entire posting."
- Thought there could be way too much data presented on the frontpage at once if the user opened all of the reviews at once.
- Hesitated She hesitated to click on the user's name to contact them as she didn't think it was a link
- Was She was surprised when the click on the username took you to the user's profile page as opposed to message sending interface. Wanted that transition to be the default behavior.
|
5 | Performed as expected. |
7 | - First jumped up to the search bar to begin the search and became frustrated that she couldn't find the advanced search options right there.
- Took a while to find the Advanced Search sidebar tab. "That isn't the last place I would look for it, but it certainly wasn't the first"
- Really liked the idea of the range searches and thought the emblematic notation was nice, but not consistent across the site
|
- After the study she as asked, "What the main purpose of your website?" We responded "To post reviews, read reviews of positions and contact others about positions." She replied "I didn't get that without your explanation. You should make it clearer on the site."
- " The emblematic design isn't really working for me. The logo for hours isn't consistent. It should be otherwise what it is representing is not easy to understand."
User 3
Task Given | Unexpected Responses |
|---|
1 | As expected, but asked "Is this the favorite button?" before clicking it. |
3 | NOTE: We performed this function incorrectly and handed the user the advanced search screen as opposed to the review screen. - The response of the user was interesting as she tried very hard to make the range functionality serve her purpose of ranking the job. She settled on showing the variability in the hours and difficulty axes, but was stumped on what to do when there for Overall Rating and Pay.
|
5 | As expected. |
6 | - Expected the unfavorited item to be immediately removed from the screen.
|
7 | - Very unclear on what "a lot" meant, but settled on only the 5 dollar signsigns. Didn't really understand how to represent this with the current expression of range. Finally ended up selecting all 5 dollar signs but was not satisfied that was the best way of presenting the information.
- Took time to fill in all of the fields for the range search(as a default they are blank)
|
- "Why can't I apply to jobs on this site?"
- While this is not the primary goal of the website,it is obvious that people will think that a job website should allow for people to apply to jobs. Perhaps this is something we should consider, however, at this time it is not one of the main goals.
Prototype V2.0
Recommended/Favorites Screen
Change in Graphs
The idea of the line graph was a poor one as we have discrete data to plot. We are now plotting the data in a line graph.
The star emblem between favoriting and overall ranking seemed to be confusing we decided to give numerical rankings to the graphs with a textual label of "Overall Ranking".
When you unfavorite something, it immediately disappears from the Favorites page.
Change in behavior on clicking on a position
Before this caused a dropdown of the reviews posted to this job, however this was deemed as too cluttered, particularly if there were multiple open. Instead if they click on the posting it causes an accordion to appear that displays bar graphs of Pay, Hours and Difficulty with a link to see reviews. There is also a link to view additional information that takes the user to the position screen that shows the textual component of the reviews in addition to the criteria of
Post Review Screen
We added Position Type fields to the top of the form. Position Type allows the users to select between a review of a full-time position and an internship position. There are now two divisions to this form: Interview Process and Position. Interview process is compressed initially(until click) and hides the rating fields: Number of Interviews, Technical versus Behavioral, Stress Level, Difficulty and Super Day.
Messages Screen
The messages screen seemed pretty spot on with what we expected. We did not make any changes to this screen.
Advanced Search
Advanced search no longer is on the sidebar. Instead it is moved "Google-style" next to the search box. Unlike Google, if you click on it an accordion drops down that allows for refining the search results for both the interview process(initially hidden in an accordion) and the position itself.
Differences to the range search, we now provide one bar per criteria(Overall, Hours, Pay and Difficulty) with two control handles that start at the extremes of the range. These handles are draggable such that the highlighted region is only the region between the two handles. If the upper bound handle is dragged to a value lower than the minimum bound(or vice versa), then the maximum and minimum move as one point.
Search Bar
We now provide suggested search completions based on what the user enters that will display in a list directly below the search bar. By selecting the one of the items in the list it will return the results of that search.
When used in conjunction with advanced search, the dropdown covers the advanced search accordion box as though it is a regular page element as it is unlikely someone will by typing and filling out that data at the same time.
Profile View
We did not construct enough test cases to effectively understand behaviors on the Profile screen(as it is not the focus of our application), but it did not cause any distress so it was left in its V1.0 state.
Prototype V2.0 Observations
...
Task Given | Unexpected Responses |
|---|
1 | As expected. |
4 | |
5 | |
| - Questioned whether it was possible to review the interview process without reviewing the position.
- Thought the extra click to open the dropdown, while easily understood, was inefficient.
|
5 | - The dropdown was a little confusing to him as it didn't separate between positions, reviews and people. Although it was in alphabetical order and she was able to quickly scan the list to find John, it actually slowed down text entry because she continued to scan the list.
|
6 | Performed as expected. | 6 | |
User 5
Task Given | Unexpected Responses |
|---|
1 | |
4 | |
5 | |
Performed as expected. |
4 | - Wanted to post sample technical questions. Ended up using the open response area to comment on it, but was worried that this might be lost. However we feel this is something that would actually be useful and important to users, particularly if they are using this to prepare for interviews.
- "It could be stackoverflow for interviews"
|
5 | - "How many 'John Curtice's are there at MIT. This doesn't tell me anything about his class year or image".
- Since we only had 2 John's in the database response it was easy to find the correct one, but he was concerned that without a full results page eventually he would have to scroll in a dropbdown box. He was not happy about this at all.
- "What if I enter 'Morgan Stanley'". He was worried that he couldn't tell the difference between the company and a person if they had similar names.
|
7 | - He tried to click in the middle of the bar. Nothing happened because that action does not indicate which of the bounds should move.
- He then clicked and dragged the Salary lower bound to the number 5. He then decided that he wanted to make both 4 and 5 as "a lot." But since he had let go of the bounded tabs they were covering each other.
| 7 | |
User 6
Task Given | Unexpected Responses |
|---|
1 | |
3 | |
5 | |
"Is favoriting the same as starring on Gmail?" |
3 | - Despite having marked the position as an internship she still included the word "internship" in the title of the position.
- We will have to make a decision on how to deal with copies of positions that share similar names
|
5 | Performed as expected. |
7 | - She used the handles correctly to select a range of payments, but didn't select a discrete value in the set {1, 2, 3, 4, 5}, instead settling between 4 and 5. Addtionally she wanted to know exactly what value she selected(i.e. 4.5) This makes us think it should either snap between the values or have a larger range like 1 to 10.
- She thought it was a little bit clumsy as the handles to the range bounds were a little small.
| 7 | |