toby
New Member
Posts: 21 
|
Post by toby on Apr 17, 2016 1:45:55 GMT
Authors:BJ Fogg: Dr. BJ Fogg directs the Persuasive Tech Lab at Stanford University. A psychologist and innovator, he devotes half of his time to industry projects. His work empowers people to think clearly about the psychology of persuasion — and then to convert those insights into real-world outcomes. BJ has created a new model of human behavior change, which guides research and design. Drawing on these principles, his students created Facebook Apps that motivated over 16 million user installations in 10 weeks. He is the author of Persuasive Technology: Using Computers to Change What We Think and Do, a book that explains how computers can motivate and influence people. BJ is also the co-editor of Mobile Persuasion, as well as Texting 4 Health. Summary:This paper presents a large-scale online questionnaire study on 1,410 web users evaluating the effect on credibility for 51 website elements. The data showed which elements boost and which elements hurt perception of web credibility. Through analysis, the authors found these elements fell into one of seven factors. In order of impact, the five types of elements that increased credibility perceptions were “real-world feel,” “ease of use,” “expertise,” “trustworthiness,” and “tailoring.” The two types of elements that hurt credibility were “commercial implications” and “amateurism.” Methods:
 In this research, the authors gathered user data through an online questionnaire shown below. The questionnaire asked the participant to evaluate 51 elements on whether they boost or hurt the perception of web credibility. 1,410 participants with average 3.7 years on the internet and 13.9 hours spent on internet per week participated in the study. The authors then ran a factor analysis on the 51 elements and determined 7 composite scales: Real-World Feel, Ease of Use, Expertise, Trustworthiness, Tailoring, Commercial Implications and Amateurism. Results: The authors presented the mean score (on a scale of -3 to +3) for each element for its impact on the web credibility. In order of impact, the five types of elements that increased credibility perceptions were “real-world feel,” “ease of use,” “expertise,” “trustworthiness,” and “tailoring.” The two types of elements that hurt credibility were “commercial implications” and “amateurism.” Based on the results, the authors gave 7 design guidelines to make a website credible: 1. Design Web sites to convey the “real world” aspect of the organization 2. Make Web sites easy to use 3. Include markers of expertise 4. Include markers of trustworthiness 5. Tailor the user experience 6. Avoid overly commercial elements on a Web site 7. Avoid the pitfalls of amateurism Thought Questions:
This study utilized online questionnaire as the research method, and asked the participant to rate 51 web elements (each is a description of a characteristic of a web site). Do you think this method is appropriate? Why? The result of this study suggests how web users judge the credibility of a website, how can you relate this paper with other readings from this week on people’s decision making process? The study was conducted in 1999. Do you think the elements that make web sites credible have changed over the time? If so, how? How is what makes web sites credible different from that of other information sources (e.g. newspapers, stories told by your friend, TV programs)? (this post was written collaboratively by Toby and Anthony) Bonus Points  For those of you filed Pittsburgh city tax, how credible do you feel about the filing website (https://www.jordantax.com/)? Why?
|
|
mkery
New Member
Posts: 22 
|
Post by mkery on Apr 17, 2016 23:43:36 GMT
The study was conducted in 1999. Do you think the elements that make web sites credible have changed over the time? If so, how?
“What makes web sites credible” has been a pretty continuous point of research after this paper. Website credibility belief is essential to e-commerce, health information, reviews, phishing websites, and cyber security. Clearly the web has changed since 1999, however, and these elements Fogg proposed need re-examining. For example: “website lists organization’s physical address” or “site lists members photos”. While useful, do we note the address of an organization that exists primarily online, like Modcloth.com or Amazon.com? Also “amateurism” and website design is an interesting point. While websites used to be quite laborious to create, the availability today of clear reusable designs like bootstrap for css or blog templates or website templates… these make it far simpler to create a professional looking (but potentially dangerous, noncredible, or scam) website. It would be interesting to repeat this study looking purely at design, to see if the evolution of easy web design has made consumers themselves more savy about how a website looks versus how much you can trust it.
|
|
|
|
Post by anhong on Apr 18, 2016 1:55:39 GMT
mkery makes a good point about the evolution of elements that makes a website credible. A lot of companies are delivery services instead of objects, and templates are making web design much easier. A quick though on the millions of webpages that uses templates, I think the level of details makes a difference. The amount of information it can present shows the credibility of the thing they are trying to sell. Also a lot of these elements expertise and trustworthiness are demonstrated through third party endorsements like verisign, norton secure, etc.
One question about the methodology of the paper is, why not asking the user comments on the credibility of the webpage, and how they feel, and why they feel about it? Then ask experts to conduct heuristic evaluations on the elements of the webpage, then try to find correlation? This can also be automated, such as scanning new online shopping and booking websites for browsing history, and analyze the drop out rate and location of the webpage, also correlate with the website's design features.
|
|
|
|
Post by sciutoalex on Apr 18, 2016 15:18:25 GMT
Medium.com recently made an announcement that a number of once-independent blogs were migrating to Medium's sparse UI. Medium sold the move as a great deal for publishers who will not have to worry about hosting a site and will have access Medium's analytics features. But the implications for what people want out of the web is clear, people want clean, well-lit websites in the same way they want clean, well-lit neighborhoods. Anyone who has studied neighborhood change knows that while that's a reasonable desire, a lot of unintended consequences come out of this desire, especially around race and class.
We can see this desire for uniformity of experience in many other places on the web. Facebook's enforced uniformity aimed at college students beat out Myspace's ethnically diverse and working class customizability comes to mind. You see it in the move away from personalized blogs to Tumblr accounts. I think this desire for a uniform experience is closely related the question of credibility in that once we have a pattern for what credibility looks like, we can apply that to every other experience that shares high uniformity.
I'm not sure I would run this study today. In 2001, people were visiting many different websites and forced to make these critical credibility judgements will little experience and context. Today, most of our time on the Internet is spent on a few services that we overwhelmingly trust (. I think if I were studying this area today, I'd want to know how people form these close, highly regular relationships with services that dominate the time spent on the computer.
|
|
|
|
Post by mrivera on Apr 18, 2016 16:01:01 GMT
"For those of you filed Pittsburgh city tax, how credible do you feel about the filing website (https://www.jordantax.com/)? Why?" I didn't file Pittsburgh city tax, because #YOLO. But when someone first sent the link to that website, it seemed incredibly untrustworthy. Apart from the (c) 2016 at the bottom of every page, the entire site seems like it was made in 2002. There are also no indications that people exist at the location (photos, names etc). The photos on the page are just of an unmarked building or random outdoor fixture. (The site desperately needs a make over. ) When i worked at FB, there was a lot of conversation revolving a logo redesign, which interestingly runs into this idea of website (more generally brand) credibility. The style, colors, font face etc all play into our perception of a particular web site. For example, banks tend to have very authoritative colors (red, blue) or colors related to money (green). Their fonts tend to be bolded or more strong (less flowy, and thin). E.g.( s-media-cache-ak0.pinimg.com/736x/7b/4c/f1/7b4cf18bf96aca44c52b51a9eb7bb460.jpg ) Anywho, to anhong's point: "I think the level of details makes a difference" - In a world of templates, the logo, and minor design aspects have a great impact (or at least more memorable effect). |
|
|
|
Post by JoselynMcD on Apr 18, 2016 21:12:46 GMT
RE: The study was conducted in 1999. Do you think the elements that make web sites credible have changed over the time? If so, how?
The study was conducted in 1999 and I think the results of that study very much reflect 1999 attitudes about what makes a website credible or not. At that time, there were far fewer websites than today, and few fewer people with the technical know-how to make their own sites from scratch (for ill or good). I think now we consider a broader range of elements, some even before we click the link, that help us decide quickly whether we trust the site or not.
Somewhat related to question one, I think that the way they studied the pages was legitimate in protocol, but today more context would have to be given. Many people use the SEO ranking as a major indicator of trustworthiness of the site. Also, people are increasingly getting to sites through social platforms like Facebook, so when their first impression of a site is far more compliated than previously, because it includes the schema associated with the person that shared the site, i.e. " This is a link to a site I've never heard of shared by Johnny, my conservative uncle who bass fishes" is going to give me a different first impression than something shared by my professor.
|
|
nhahn
New Member
Posts: 22
|
Post by nhahn on Apr 18, 2016 23:45:37 GMT
In some of the work I was recently doing, I had to figure out the best way to display an article that I had a bunch of crowdworkers write together. The crowdworkers had to use snippets of information collected from the web to produce and article about a specific topic. We were curious if we changed the design of the website around to look like wikipedia (compared to our original design, which was slightly more optimized from the content), if people would trust it more. We ended up having significant results, where the perceived trustworthiness / credibility of the article was higher when it was modeled to look like wikipedia. There are several possible explanations. One could be that the wiki-format is the best way to present information aggregated / worked on by a number of different individuals. Another could be individuals thought it was something that was from wikipedia, so it was inherently more trustworthy. Yet another could be this match between the content type and the way the article was written. Do you think this is an easy way to trick people into believing information that could very well be false? What do you think was in play that caused this change in trustworthiness perception?
|
|
|
|
Post by Nick on Apr 19, 2016 0:22:00 GMT
What do people think about the "Trusted Website" or "Secure" badges (for example: www.copyrightsworld.com/images/logos/secure.jpg)? I distinctly remember an instance where I had the thought, "Is this website secure...oh, well it has the paypal certified badge, it can't be too bad CLICK" and then immediately thought, "That was just an image Nick, you should not put so must trust in an jpeg." Despite knowing better, that badge still created some credibility for me. To MaryBeth's point, I wonder how this has evolved over time not just in terms of website elements, but just because the landscape of the internet has changed. I think it's certainly more important to know if a website is trustworthy today than in 1999, just because so much more of our lives are spent online. Which makes you wonder how much the "real world" aspect of an organization is important. Is twitter's "real world aspect" important? When Amazon opened their first brick and mortar store, it was their "digital world" aspect that was important. I think the prevalence of the internet and our investment in it are blurring that line. |
|
|
|
Post by stdang on Apr 19, 2016 0:32:09 GMT
How is what makes web sites credible different from that of other information sources (e.g. newspapers, stories told by your friend, TV programs)?
While I think the general values that any individual applies to evaluating the credibility of a news source are likely similar across news sources. I think that the cues that any person uses to assess credibility along any particular dimension for a given class of news source (eg: Newspaper, online blog, online news paper, tv) are going to be different and related to the affordances of communicating via that particular medium. To this extent, I don't think the general notions of credibility as applied to websites have changed since 1999. However, because the culture of the internet has changed the cues that signal credibility have likely changed. There are likely specifics in terms of page layout, color schemes, and domain specific language that are all indicative of credibility for online websites. While these may have been the same as in 1999 (in order to assess the 7 criterion discussed in the paper), the thresholds for each of these is likely higher as standards on a whole have risen since the 90's.
|
|
|
|
Post by Anna on Apr 19, 2016 0:41:02 GMT
Hm, Alex brings up some important points about web uniformity, classism, etc. I also think Joselyn's point about context mattering is very relevant. I found it interesting that the authors didn't ask participants to distinguish between different sites in any way (or did I miss something??). A lot of the criteria simply wouldn't apply to different types of sites (e.g. news stories, awards, outside materials and sources, etc.). I have totally different sets of standards for different websites. Certain elements-- especially aesthetic elements-- that might make one site feel untrustworthy become irrelevant in another site. To take the tax example, the fact that the jordantax site looks like it was made in 1993 makes me shake my head at Pittsburgh a little, but it doesn't actually make the site less trustworthy to me-- I still feel totally comfortable submitting my taxes there. However, if I came to a new shopping website that looked like that, I would at least think twice before submitting my credit card info. In some situations, I may even be more suspicious of a more polished site than a 90s-looking site. Take this forum site, for example, or really any online forum community-- I find the lack of attention to aesthetics leads to a pleasantly gritty user experience, and I sometimes assume that the conversation is more likely to be more authentic, or more serious. Similarly, if I go to a store or restaurant website and it's straight out of the 90s, I might actually have higher opinions/expectations, because I might think that it's a mom-and-pop place with a heavy focus on the in-store/in-restaurant customer experience.
|
|
|
|
Post by francesx on Apr 19, 2016 1:22:16 GMT
Re: This study utilized online questionnaire as the research method, and asked the participant to rate 51 web elements (each is a description of a characteristic of a web site). Do you think this method is appropriate? Why?
I am always a little bit skeptical when it comes to questionnaires and liker scales. Doesn't everyone have their own definition of what each of these numbers means? On the other hand, I cannot think of any other way or method to do the same task (maybe some sort of design method that aims to explore and find tacit knowledge?)
Re: The study was conducted in 1999. Do you think the elements that make web sites credible have changed over the time? If so, how?
Certainly some elements have changed, but I do agree still with a lot of points mentioned in the study (include markers of expertise, include markers of trustworthiness, avoid overly commercial elements on a Web site, etc.) For example, when I look into websites in other languages (excluding English) I generally see more of those guidelines not being applied.
|
|
|
|
Post by Amy on Apr 19, 2016 2:49:56 GMT
Taking Alex and Jocelyn's points together - a lot of the sites we visit are already established as credible, or are referenced to us from credible sources. So how much "evaluation of credibility" does the average user really do? Is website credibility the right question to be asking?
I think a way to modernize this study is to look at what makes people decide an app is credible enough to download. I think there are more pertinent security and trustworthiness issues there, as we have had a lot more "training" on how to evaluate websites than how to evaluate apps.
|
|
judy
New Member
Posts: 22
|
Post by judy on Apr 19, 2016 3:09:48 GMT
Alex, danah boyd writes about the difference between myspace and facebook (when both of them existed at once) and how teens expressed cultural identity and socio-economic status through being on one platform or the other (http://www.danah.org/papers/2009/WhiteFlightDraft3.pdf and the book: www.danah.org/itscomplicated/). White, middle class teens though myspace was "ghetto," and they could see school culture split down myspace vs. facebook lines. Our ideas about credibility have shifted. For example, when I started teaching high school in 2006, librarians everywhere were trying desperately to get a handle on how to teach research in the age of the internet. Using wikipedia as a source was strictly forbidden. And all freshman had to take a mini-course on how to recognize a "credible website" (stuff like university affiliation). As Nathan's study points out, now we see wikipedia as a trustworthy source. I recently did a lit review on credibility in social media. And it's funny, over the last ten years, credibility has basically collapsed to the analytics--how many likes something has, whether it's trending, where it falls in a google search. Is it good to put so much emphasis on what's popular? What messages or truths get left out if that's how we define credibility? |
|
|
|
Post by jseering on Apr 19, 2016 3:19:31 GMT
Per the collapse of credibility in Judy's post- are we surprised that we're focused on measuring the things that are relatively easy to measure? Ideally we wouldn't let what's measurable define what questions we ask, but we have deadlines to meet. Sorry, it's been that kind of day.
Cynicism aside, this conversation reminds me about Donath's work on social signaling. We can assume a human is credible online when they're able to produce signals that are reliable markers of their real status. Posting on a forum that you're Oprah isn't a particularly reliable signal, but posting from a verified twitter account probably is. One of the challenges with anonymity is that it's hard to know who to trust; it's difficult to evaluate the expertise and honesty of the people posting or sharing. This is especially the case where pseudonyms aren't used and users can't establish credibility over time. One of the interesting challenges in designing anonymous apps is creating a positive community without the signals that we usually rely on to assess each other's behavior.
|
|
|
|
Post by fannie on Apr 19, 2016 3:32:26 GMT
I like Alex’s point about doing a study about people’s relationships with certain services - it reminds me of the people’s mental model of the internet that we saw in an ARM thought question where someone literally just drew an Earth and Facebook. I’m also curious about how credibility might change with the integration of these services with other websites (log in with your Facebook account, Twitter feed sidebars). Popularity and general social norms surrounding web services are probably another factor that’s pretty prominent - I often go to websites as a result of someone sending me a link or telling me about some service. Same with restaurants, these days I rarely walk into a restaurant unless I’ve checked it on Yelp and see a sufficient number of stars.
In regards to Anna’s comment - I’ve also had similar experiences I might find less polished/modern-looking websites (or stores) more credible. For instance, when I look at a professor’s personal website (at least in CS), I often expect white or pastel color backgrounds with lots of Times New Roman text and standard blue links and a picture of them from the 90s, and when I see a modern, nicely designed website with minimal text I might be initially confused. For me, there’s some kind of schema there that may help me determine what markers of expertise or “real world” might be.
|
|
?
