What Variables Affect Web Credibility? A Report of Ongoing Research from Stanford’s Persuasive Technology Lab BJ Fogg (bjfogg@stanford.edu) First presented at CSLI’s Industrial Affiliate Conference Stanford University, November 11, 1999

Overview Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Level 1 A First Look at Credibility Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Why does Web credibility matter? Web sites with optimal levels of credibility Creators What to believe on the web (aka “information quality”) Consumers A key factor in Web-based persuasion PT Lab Answer #1

Web sites with optimal levels of credibility

What to believe

on the web

(aka “information quality”)

A key factor in

Web-based persuasion

Why does Web credibility matter? Few have studied this (publicly) Web has least credible info of any medium Web has most credible info of any medium Web standards are taking shape now The Web is not going away any time soon Big money is at stake Web presence is vital for many companies Explosion of Web-based info & services Web users are vulnerable Answer #2

What is “credibility”? Believability or A perception based on two factors* . . . trustworthiness expertise honest truthful unbiased powerful intelligent knowledgeable experienced good *Some studies have shown three or more factors +

Believability

or

A perception based on two factors* . . .

Highly credible web sites have. . . High levels of perceived trustworthiness and expertise perceived trustworthiness perceived expertise perceived credibility +

High levels of perceived trustworthiness and expertise

One factor can damage credibility perceived lack of trustworthiness perceived expertise perceived lack of credibility + perceived trustworthiness perceived lack of expertise perceived lack of credibility +

Some semantic issues “Credibility” and “Trust” Similar terms but not the same construct Phrases that refer to credibility “trust the information” “accept the advice” “believe the output”

“Credibility” and “Trust”

Similar terms but not the same construct

Phrases that refer to credibility

“trust the information”

“accept the advice”

“believe the output”

Expertise & Trustworthiness In these examples, what elements convey expertise? What elements convey trustworthiness?

Expertise & Trustworthiness In these examples, what elements convey expertise? What elements convey trustworthiness?

Level 2 Four Types of Credibility Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Four Types of Credibility 1. Presumed Credibility 2. Reputed Credibility 3. Surface Credibility 4. Experienced Credibility It’s helpful to distinguish different types of credibility. But you won’t find these distinctions in the psychology literature.

1. Presumed Credibility

2. Reputed Credibility

3. Surface Credibility

4. Experienced Credibility

It’s helpful to distinguish different types of credibility.

But you won’t find these distinctions in the psychology literature.

1. Presumed Credibility We believe because of general assumptions we hold.

We believe because of general assumptions we hold.

1. Presumed Credibility We believe because of general assumptions we hold. Examples* . . . Increases credibility Decreases credibility Domain name has “.org” Lots of hits on web counter Info on the site is constantly updated Site has AOL domain name Very few hits on web counter Info on the site is rarely updated *These are working hypotheses. To date, we’ve studied only some of these variables.

We believe because of general assumptions we hold.

Examples* . . .

*These are working hypotheses. To date, we’ve studied only some of these variables.

2. Reputed credibility We believe because of a reference from a third party .

We believe because of a reference from a third party .

2. Reputed credibility We believe because of a reference from a third party . Examples* . . . Increases credibility Decreases credibility Your medical doctor referred you to this Web site. The site won an award. An authoritative Web site linked to this site. Your friend said the site was horrible. The newspaper said the site was down for three days. A political group you don’t like endorses the site. *These are working hypotheses. To date, we’ve studied only some of these variables.

We believe because of a reference from a third party .

Examples* . . .

*These are working hypotheses. To date, we’ve studied only some of these variables.

3. Surface credibility We believe because of what we find on simple inspection .

We believe because of what we find on simple inspection .

3. Surface credibility We believe because of what we find on simple inspection . Examples* . . . Increases credibility Decreases credibility Site looks professional. Site is from an organization you recognize. You see that articles have citations. Site looks confusing. The organization has no presence outside the Web. The site uses many animated features. *These are working hypotheses. To date, we’ve studied only some of these variables.

We believe because of what we find on simple inspection .

Examples* . . .

*These are working hypotheses. To date, we’ve studied only some of these variables.

4. Experienced credibility We believe because of past experience with site.

We believe because of past experience with site.

4. Experienced credibility We believe because of past experience with site. Examples* . . . Increases credibility Decreases credibility You get a quick response to a customer service question. You can navigate the site easily. You’ve found the content to be fair and balanced. The site has a broken link. The site takes a long time to download each page. You’ve seen factual errors on the site. *These are working hypotheses. To date, we’ve studied only some of these variables.

We believe because of past experience with site.

Examples* . . .

*These are working hypotheses. To date, we’ve studied only some of these variables.

Why this framework is helpful Provides a set of terms for discussion Enhances both research & design Asking better questions Seeing more possibilities Presumed Credibility Reputed Credibility Surface Credibility Experienced Credibility Trustworthiness Expertise

Provides a set of terms for discussion

Enhances both research & design

Asking better questions

Seeing more possibilities

Why this framework is helpful Provides a set of terms for discussion Enhances both research & design Asking better questions Seeing more possibilities Presumed Credibility Reputed Credibility Surface Credibility Experienced Credibility Trustworthiness Expertise 

Provides a set of terms for discussion

Enhances both research & design

Asking better questions

Seeing more possibilities

Four Types of Credibility Presumed Reputed Surface Experienced Can you identify the different types of credibility in these examples?

Presumed

Reputed

Surface

Experienced

Four Types of Credibility Presumed Reputed Surface Experienced Can you identify the different types of credibility in these examples?

Presumed

Reputed

Surface

Experienced

Level 3 Web Site Variables Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Level 3 Web Site Variables How to contain the many elements of a Web site? We’ve created some frameworks I’ll talk about one. See www.webcredibility.org for others.

How to contain the many elements of a Web site?

We’ve created some frameworks

I’ll talk about one.

See www.webcredibility.org for others.

Framework for Web Site Elements Web Site Provider Person or organization that offers the site e.g., IBM, Kodak, Amazon Web Site Content Information Functionality Web site Design Aesthetic design Information design Technical design

Web Site Provider

Person or organization that offers the site

e.g., IBM, Kodak, Amazon

Web Site Content

Information

Functionality

Web site Design

Aesthetic design

Information design

Technical design

Web Credibility Grid Web Site Provider Web Site Content Web Site Design Presumed Credibility Reputed Credibility Surface Credibility Experienced Credibility information functionality aesthetic information technical

information

functionality

aesthetic

information

technical

Web Credibility Grid Web Site Provider Web Site Content Web Site Design Presumed Credibility Reputed Credibility Surface Credibility Experienced Credibility The provider is a nonprofit organization. The site was created by an outside design firm. The site has ads from reputable companies. Users are familiar with the provider outside of the Web context. The provider is recognized as an expert by others. The site’s content has always been accurate and unbiased. The site is easy to navigate. The site appears to have lots of relevant information. The content has been approved by an outside agency. The site won an award for technical achievement. The site has a pleasing visual design. Users with questions receive quick and helpful answers. Examples of elements that increase credibility* information functionality aesthetic information technical *These are working hypotheses. To date, we’ve studied only some of these variables.

information

functionality

aesthetic

information

technical

*These are working hypotheses. To date, we’ve studied only some of these variables.

Web Credibility Grid Web Site Provider Web Site Content Web Site Design Presumed Credibility Reputed Credibility Surface Credibility Experienced Credibility The site tries to recruit advertisers but has none so far. The site has no security protocols for transactions. The site shows only a few hits on their web counter. The site’s URL does not match the provider’s name. The provider was sued for patent infringement and lost. The site has typographical errors. The site has links to pages that no longer exist. The site seems to have more ads than information. The content got bad reviews from an outside agency. The site is reported to have copied the design of another site. The text font is either too large or small to read comfortably. The site doesn’t give contact information anywhere. Examples of elements that decrease credibility* aesthetic information technical information functionality *These are working hypotheses. To date, we’ve studied only some of these variables.

aesthetic

information

technical

information

functionality

*These are working hypotheses. To date, we’ve studied only some of these variables.

Small Survey on Web Credibility Pilot study on 45 variables Questionnaire, n = 25, with Nina Kim Purpose: To select variables for experiment Sample question: The Web site lists an author for each article. -3 -2 -1 0 +1 +2 +3 much much less more credible credible

Pilot study on 45 variables

Questionnaire, n = 25, with Nina Kim

Purpose: To select variables for experiment

Sample question:

The Web site lists an author for each article.

-3 -2 -1 0 +1 +2 +3

much much

less more

credible credible

Survey results were interesting * more credible less credible Was referred to you by a good friend Contains a link to a site you trust Contains an ad for IBM Contains many spelling and grammatical errors Was created with PageMill or FrontPage software Charges a monthly access fee Contains broken links Contains misleading ads Makes it difficult to distinguish ads from content Contains a typographical error Has “.com” in the domain Contains animation Has “.org” in the domain Lists an author for each article Lists a physical address for the page sponsor Is easy to navigate Provides quick response to a customer service question Presents information that may help its competitors Represents a company or organization you recognize *The sample size in this survey was too small to draw firm conclusions

Lab Experiment on Web Credibility Selected two Web credibility variables for Nina Kim’s thesis experiment ads typos

Selected two Web credibility variables for Nina Kim’s thesis experiment

ads

typos

Web credibility research Experimental Design Controlled lab study 2X2 mixed design, n = 38 Health information taken from Web Independent variables typos or no typos ads or no ads Dependent variables questionnaire after viewing

Controlled lab study

2X2 mixed design, n = 38

Health information taken from Web

Independent variables

typos or no typos

ads or no ads

Dependent variables

questionnaire after viewing

Web credibility research Sample Stimuli No typos Typos vs

Web credibility research Sample Stimuli No Ad Ad vs

Web credibility research Sample Stimuli Other sample pages

Web credibility research What did the data show? Typos decreased the credibility of page. Ads increased the credibility of page. Both effects are stronger for vital info than for less vital info.

Typos decreased the credibility of page.

Ads increased the credibility of page.

Both effects are stronger for vital info than for less vital info.

Level 4 Variables Outside the Web Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Level 4: Variables Outside the Web Individual variables user differences (age, experience, personality, etc.) Goal variables For example, information seeking on health surveillance of current events passing time, entertainment Context variables setting time factors technology factors

Individual variables

user differences (age, experience, personality, etc.)

Goal variables

For example,

information seeking on health

surveillance of current events

passing time, entertainment

Context variables

setting

time factors

technology factors

Level 5 Three Moving Targets Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

Level 5: Three Moving Targets Three things are changing . . . 1. People Web user base will become more diversified. 2. Experience levels Web users will become more savvy. 3. Technology Web technology and uses will change. These moving targets mean . . . Perceptions of Web credibility will shift over time. We need an ongoing effort to understand Web credibility.

Three things are changing . . .

1. People

Web user base will become more diversified.

2. Experience levels

Web users will become more savvy.

3. Technology

Web technology and uses will change.

These moving targets mean . . .

Perceptions of Web credibility will shift over time.

We need an ongoing effort to understand Web credibility.

Ongoing research efforts at PTL Research on web credibility perceptions www.webresearch.org Web credibility experiments Online and in the lab Gathering other work on Web credibility Pointers at www.webcredibility.org

Research on web credibility perceptions

www.webresearch.org

Web credibility experiments

Online and in the lab

Gathering other work on Web credibility

Pointers at www.webcredibility.org

In Review Level 1 A First Look at Credibility definitions, basic issues Level 2 Four Types of Credibility different kinds of credibility Level 3 Web Site Variables web elements and a framework Level 4 Variables Outside the Web individuals, goals, and contexts Level 5 Three Moving Targets people, experience levels, and technology Plus two studies

Level 1 A First Look at Credibility

definitions, basic issues

Level 2 Four Types of Credibility

different kinds of credibility

Level 3 Web Site Variables

web elements and a framework

Level 4 Variables Outside the Web

individuals, goals, and contexts

Level 5 Three Moving Targets

people, experience levels, and technology

What Variables Affect Web Credibility? A Report of Ongoing Research from Stanford’s Persuasive Technology Lab BJ Fogg (bjfogg@stanford.edu) First presented at CSLI’s Industrial Affiliate Conference Stanford University, November 11, 1999