of two communication products and compare outcomes. Option 2 Test usability of one website. OR Test usability of two websites and compare outcomes. 2 The next few slides will give you some ideas for what to test based on each option. You may use any of these ideas or come up with your own.
Open Office, Google Docs Image Editing Gimp, Adobe Photoshop, Pixelmator Website Creators Wix, Weebly, JIMDO 3-D Modeling AutoCad, OpenScaD, SolidWorks, FreeCAD MOBILE APPLICATIONS You may test two apps. These may be the mobile versions of software or stand-alone Apps. Just be certain any stand-alone apps are complex enough to test USABILITY. Ask me if you are uncertain.
They don’t need to have the exact same features. But they need to be the same type of product. Example: AutoCad does not have the same features as SolidWorks but both are 3-D modeling tools, and Pixelmator does not have the same features as Adobe Photoshop but both are image creators/editors.
choose to test one website or compare two websites that are similar such as two department of fish and game sites or two university sites. The key is to choose a website that has complexity, where users need to find and/or do stuff. • I provide the option of choosing just one website because often it is not possible to compare two sites if what you want to test is not accessible to you. • For example, you culd not compare how to enroll in classes at the University of Idaho compared to Boise State because you would not have access to both of these.
comparable. 7 They don’t need to have the exact same features. But they need to be the same type of product. Example: Some states have fish and game websites with more features than others, but both are the same type of website. You may also compare two university sites from the viewpoint of a user who is not enrolled in either: a prospective student. x
product or website and decide their level of knowledge/background. For example, engineering students who have never used AutoCad or SolidWorks would bring a level of knowledge when using the products that others outside of engineering would not possess. A hunter using the Washington Department of Fish and Wildlife website would bring a level of knowledge to the site that a non-hunter would not possess.
products or two websites, you will need to test approximately 10 specific tasks. Avoid tasks that are too basic to evaluate usability. Example: testing how to save a document file. • If you are testing one website, you will need approximately 20 specific tasks (refer to slide 6 in this slidedoc if you are confused about why I offer the option of testing just one website). • If you have no idea what type of tasks can be done with a product, then consult the product’s HelpCenter only to determine what types of things the product or website will let users do.
screenshot of Weebly’s Help Center. Do not spend time trying to learn how to do a task because you will be testing learnability as one component of your test. Simply use the help page menus to derive a list of the tasks.
which you will evaluate the following five components that define usability. 1. Effective: were you able to complete the task? Also means achieving goals. 2. Efficient: how long (or number of clicks) did it take you to perform a task and is that amount of time reasonable? 3. Engaging: Did the product’s style and appearance enhance your user experience? 4. Error Tolerant: How many times did you attempt an action that did not work? when you made an error or a false start, did the product give you feedback where needed? 5. Easy to Learn: how easy was it to learn how to do a task?
need to include the following: – Qualitative Data: five criteria statements aimed at evaluating the five components that define usability. You must write STATEMENTS and not questions. – Quantitative Data: a fixed-response Likert Scale that allows you to measure your level of agreement or disagreement with each criteria statement.
responses to a survey. He showed how asking respondents to place themselves on a scale of favor/disfavor with a neutral midpoint worked better than other means of measuring responses. The Likert scale has since been adopted throughout the world. 2.3: Design Test Metrics Continued About Likert Scales
the use of a 5-point scale, researchers have since argued for more points to increase the reliability and validity of the scale. Their arguments are based on findings from studies about how respondents choose items on the scale. 2.3: Design Test Metrics Continued
• Respondents may be unmotivated to make an effort to choose anything else. • Respondents may be ambivalent. • Respondents may be reluctant to voice a socially undesirable response. 18 Based on these findings, some researchers advocate eliminating the midpoint or adding more items to the scale. 2.3: Design Test Metrics Continued
Only: respondents are more attracted to labeled points, so just labeling the end points may result in a bias towards extreme answers. • Labeling All Points: respondents are more attracted to intermediate options, so it can lower extremeness of responses, which is good. But, it may also increase levels of positivity bias, the tendency to respond with a positive answer, which is not good. 21
experiments on short-term memory. • He published the results in a paper: The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information. • His experiments lead him to conclude we can only retain 5 to 9 items in our short term memory. 24 It’s your choice, but here are some things to think about.
for all criteria statements. • Don’t switch from a five-point to a seven-point and then to a four-point. • Keep the positions of the value labels the same. • If you start with negative type value labels on the left of the scale (levels of disagree) and positive on the right (levels of agree), don’t change the positions. • Example of inconsistent value labels:
has to be written in a way that a respondent (in this project the respondent is you) can answer by rating the level of agreement with that statement. • Ease of use. • I was able to complete the task without consulting the product’s HelpSection.
for each task you do. • You may create a data collection form for this purpose or use a spreadsheet. • Your form or spreadsheet will need to list each tasks with the five criteria statements plus the Likert Scale and a place to take notes. 30 2.4: Create Data collection Form
white paper write-up of test findings, you will need to describe the test environment (where you conducted the test) and the equipment you used (PC? Mac? Operating system? Etc.). • The test environment should be a quiet place where you can work without interruption. • If you do the test in more than one setting, be consistent with the environment and equipment you use.
what happens when you do a task in the five component areas that define usability. • Score your level of agreement or disagreement with the criteria statements in the data collection form or spreadsheet. AND • Take notes to help you describe what happened as you did the task and to offer any recommendations about how to improve the product or website.
notes. • The number of screenshots you take should work to support the descriptions you will write. • You will need to size these to be appropriate to the page layout in your white paper. • Often screenshots are difficult to wrap text around. That’s fine, but do keep the screenshot close to the textual information it is supporting. Remember to label, caption, and refer t screenshots in the text.