In the context of News and News media, the new gadget iPad should not be ignored. Dynament is not especially designed for the iPad device. There were some points obviously not working, such as the small navigation links in the public container or the small interaction icons in the private container, which were designed for mouseinteractions and not touch. The tests worked well on iPad. All the interactions, although not designed for touch, worked in their functional purpose, but did not feel as intuitive as could be. I conclude that if a device such as iPad is considered in the concept, it must be included in a very early stage of design development. There are many mouseinteractions which just are not the same for touch as for the mousepointer.
In the context of News and News media, the new gadget iPad should not be ignored. Dynament is not especially designed for the iPad device. There were some points obviously not working, such as the small navigation links in the public container or the small interaction icons in the private container, which were designed for mouseinteractions and not touch. The tests worked well on iPad. All the interactions, although not designed for touch, worked in their functional purpose, but did not feel as intuitive as could be.
I conclude that if a device such as iPad is considered in the concept, it must be included in a very early stage of design development. There are many mouseinteractions which just are not the same for touch as for the mousepointer.
Iterated Prototype
Published May 28th, 2010 in 05 Prototype Development, 05.4 Iterated Prototype and Video. Closed |
Here a list of what was changed between Dynament Beta 1.0 and Dynament Beta 2.0
- Public Navigation: The public navigation to browse the comments now distinguishes between comments and discussions. Discussions are now having a overline over the number. The navigation has as well been enlarged. The navigation control is now only shown once and not repeated for all comments. At the moment the most actual three comments are shown (instead of only one)
- Page Reload: The page reload in Beta 2.0 is extended with a functionality that jumps right back to the paragraph where the user made the comment. Not implemented is that the comment should as well be shown immediately.
- Replying on a paragraph vs. replying on a comment: To make it more clear Beta 2.0 features two different icons: one for commenting on a paragraph and another for commenting on a comment.
- Discussion and orientation: Beta 2.0 highlights on rollover the users involved in a discussion which makes it easier to follow up discussion threads and see the inherent relationship.
- All Comments and reference to paragraph: At the end of each article the user has the possibility to show all comments belonging to the article. The comments which were originally written to a paragraph feature a link which jumps directly to the paragraph where the comment has originally been posted.
- General Comments: Beta 2.0 includes the possibility to make general comments to articles. These comments are listed as usual at the bottom of the article.
- Embedding into NZZ Online skin: Beta 2.0 is embedded in a fake NZZ-Online environment. Most NZZ-links are not available but add up to the appeareance and the general feeling.Setting up several articles environment: Beta 2.0 features 5 different articles, so there is a certain variety of articles users can choose from. As for the exhibition the articles were differently populated with comments (from none to many). The exhibition is therefore another kind of user test in regard of response volume of comments.
First Prototype ‘User Tests 01′
Published May 10th, 2010 in 05 Prototype Development, 05.2 First Prototype and 05.3 User Tests. ClosedThe prototype was tested on 15 Users. Since Dynament addresses mostly computer affine people, I tested it on 10 users which met this criteria. This in order to get a feedback from the main user group. The other 5 probands were people with average computer skills, from whom I hoped to gain insight into problems which more advanced users generally avoid.
The 10 advanced users were only given little information about the project. They were asked to read a highly controverse, polarizing article about the flyer of the BA-exhibition. They were additionally asked to comment what they read.
The other 5 probands were given an additional introduction into the concept and were shown certain interactions beforehand.
After the test which each lasted about 5 minutes there was a talk about the experience, weak points, and wishes of users.
| Successful User Test: | Unsuccessful User Test: | |
User Behaviour
Same as in the Cognitive Walkthrough (see: 3.3 Cognitive Walkthrough) the probands displayed individual beahavioural features interacting with Dynament. Namely the group of advanced users tended to comment on the go, whereas the unexperienced users preferred to first read the article and then place comments.
None of the probands actively perceived the division into private, text and public parts, but almost all of them recognized it as evident and good solved when hinting at the division, in the talks afterwards.
A majority of the probands wanted to see more than just one comment at the same time, but not all at once. Two did not succeed in navigating further than the first comment.
Four probands were confused with the difference between commenting on paragraphs and commenting on others users comments (discussing).
Six experienced it as exhausting to comment on paragraphs; four of them had very well an opinion but did not find a suitable paragraph to place it, the other two just did not care about accuracy and posted a comment anywhere.
One proband deleted by mistake the text-input and sent an empty comment.
One proband got confused when more than one text-input box was visible on the screen.
All probands who succeeded in posting a comment found it confusing that the page reloaded without displaying, or at least highlighting their comment.
Four probands were dissatisfied by the discussion depth, it was not clear to them who commented on whose comment and so on.
When asked seven probands indicated to have enjoyed the experience and propagated Dynament an additional value. Four were not convinced by the concept and preferred to not comment at all or to make general comments at the end of the article. Two probands could understand the additional value but did not like to comment with paragraph-accuracy and two did not have a final opinion.
Users Inputs and Wishes
The probands wishes and inputs were collected and each of it proposed to the next proband. It turned out that many users wished to have easier access to the comments and know more about the type of comment (wheter it was a single comment or a discussion). They wanted to see more than just one comment at the time, but not all at once. One suggestion was to implement a rating system and initially only show the top rated comments. The idea of the rating system was very well received by the next probands.
Almost all probands found it difficult not to have the possibility to place general comments at the end of an article and desired to have at least the option to do so. A further suggestion of a proband was to include social media in some kind of form.
Another well received suggestion was the wish for a widget where all the comments for an article are listed, containing hyperreferences to the main article.
Analysis and Finding
What worked or did not work as expected?
As expected the core functionality of Dynament, the commenting and reading of comments worked out pretty well, which was not that big of a surprise, since it oriented itself mainly on common, well-known conventions.
The public navigation part works in its functional purpose but is considered as not that well designed. The main problems here are: Display of more than one comment, small size of navigation, the repeating of the whole navigation for every comment, missing information about the kind of comment (comment or discussion).
Another problem is the rerendering of the page after a comment has been committed. There must be defenitely more vidual dynamic guidance to the users interactions with the comments.
As for the visual guidance in discussions it seems defenitely not to be enough to just indicate who responded to whom by simply displaying the usernames. Particularly in long discussions one easily loses the overview.
Further there should defenitely be a possibility to place general comments as well as general ratings for an article.
What worked or did not work unexpectedly?
Not that predictable for me was the problem the users had with the distinction between commenting on paragrpahs and commenting on comments (discussing). The tool distincts mainly by the color: If the interaction takes place on the public layer the color is orange, whereas if the interaction is performed in the private section it is blue. When probands tried to reply on a comment, they often commented again on the paragraph, which against the concept, did not lead to a discussion.
Another point which was unexpectedly unclear to the probands, was the embeddment of Dynament into a news and media context, which will hopefully be resolved when embedding the iterated prototype in the NZZ-Online format.
First Prototype: Dynament Beta 1.0
Published May 3rd, 2010 in 05 Prototype Development and 05.2 First Prototype. ClosedThe first prototype for the project entitled Dynament Beta 1.0 is technically realised with ruby on rails. It features Comments on paragraphs, and comments on comments (discussions). Visually it is based on the sketches of Sketches 03. This prototype was presented in the context of the third BA-presentation, where it gained valuable critique from the audience.
Main points of critique and advice were:
1) Embedding prototype into a real-world context
2) Take devices such as iPad into consideration
3) Vagueness about comment-indications and number of comments
Since Beta 1.0 was presented by myself, I could easily prevent bugs and interaction-procedure flaws. The next step was therefor to make user tests with Dynament Beta 1.0, in order to gain information about user behaviour and yet undiscovered program bugs.
Here some screenshots of the first Prototype.
