2. Summary
• Use of the Internet to look for health information
• Why credibility matters?
• Theories about credibility and Internet credibility
• Methodology and sample
• Results and discussion
3. Health information and the Internet
38% of European Population, 2011
Iceland, 61%
Finland, 58%
Denmark, Germany, 54%
Spain: 38%
UK: 36%
Source: Eurostat
4. Health information and the Internet
Changing paradigm
‘Informed patient’ (Kivits 2004):
Challenging authority of doctors
Helps in dealing with the illness
Doctors:
Same criteria as traditional medical sources
(recommendations)
Uncertainties: huge amount of data,
disparate quality, disorganization,
inadequacy of content, who to trust
5. How do users decide?
Credibility
Conceptual tool
Select
Assess
Research Precedents:
Aristotle
Mass Communication Research
(1950s)
New interest after Internet
popularization since 1990
6. Theoretical frameworks
1. Information literacy: a checklist
2. Unifying Framework of Credibility Assessment (Hilligoss y Rieh,
2008)
3. Prominence-Interpretation Theory (BJ. Fogg, 2003)
4. Cognitive Authority Theory (Patrick Wilson 1983, Reijo Savolainen,
2007, Soo Young Rieh, 2008)
5. Other studies: Rieh and Danielson (2007), Petty and Cacioppo
(1986), Metzger (2007), Wathen and Burkell (2002), Fritch and
Cromwell (2001)
7. Unifying Framework of Credibility
Assessment
Multidimensional process in 3 levels :
Construct
Truthfulness, expertise, trustworthiness...
Heuristics
Content, media, source, reputation, coherence...
Interaction
Verification-contrasting, source evaluation,
message clues
Context as a cross factor
8. Cognitive Authority Theory
Social epistemology model
Knowledge:
First hand: better but limited
Second hand: initial suspicion, the source needs the recognition as
an ‘authority’ to be trusted
Cognitive authority:
The assessment is done by “indirect tests or indexes of credibility”
(Wilson, 1983)
Facets/aspects: trustworthiness, truthfulness, scholarship,
credibility, officiality and competence
What the cognitive authority says is “self knowledge”
9. Limitations
Inaccuracy in the definition of the concept
Ambiguity believability/credibility
Very few explanations for Internet consumption
10. Credibility: an operational definition
Credibility is the quality of truth for information
not directly witnessed by a receiver that is
assessed through a series of attributes.
11. Methodology
Navigational interview
Semi-structured
It’s hold in front of a computer
The informant can explain and show what they actually do (think aloud
protocol)
Results: alows to gather explanations and do observation
Sample
Doctors (familiy doctors and hospital doctors)
Patients and relatives
Fieldwork : November 2011-February 2012
12. Hypotheses
H1 dependence on the context
H2 dependence on verification
H3 dependence on time
H4 appearance of levelling
13. Results
Doctors
Construction: complex concept with equal importance of the
attributes (expertise, trustworthiness, believability)
P20. Credibility is related to truthfulness and the truth must be
able to be checked, from a medical point of view, through what we
call evidence-based medicine.
14. Results
P11. If you read an article in the British Medical Journal or similar
journals, you already know that they have a code of ethics, that
there isn’t a conflict of interest, etc.. and, for me, this is a source of
credibility.
15. Results
Doctors
Heuristics:
Reputed medical journals with impact factor (expertise)
Oficial health sites (expertise)
Medical associations (expertise)
Blogs or webs by colleagues (trustworthiness)
Message signs: methodology description (expertise)
Updated content (expertise)
Comparison among different sources (believability)
Reminder for specific information-clinical practice (believability)
16. Results
Doctors
Interaction:
P20. Many times, and although it contradicts what I’ve told you
before, I use Google because we cannot disregard its results (...)
many times it brings you to the same web places.
Scientific databases and journals
They use of Google (means of “transportation”, look for fast
information...)
They look for the description of the source
17. Results
Patients
Motivation: to know more about their medical condition, dealing
with the illness day by day (side questions), learning about
scientific advances
Concept: expertise (medical-research credentials)
P15. I rather trust the opinion of a researcher... Maybe I trust more a
researcher than a doctor because they have more expertise in a
particular topic.
18. Results
Patients
Concept: expertise-trust-believability are strongly related
P16. The expert deserves my confidence, but not absolutely. If
someone comes here and says ‘I am the king of cystic fibrosis
research, I am doctor X...” I thoroughly observe him and I do not
trust him blindly. First I have to trust. So, where does the trust
come from? From a previous relationship. Expertise is important
but must be combined with trust and plausibility, too.
19. Results
Patients
Concept: general scepticism on the Internet (just one case).
P17. In my view a doctor speaking on the TV or on
the radio has more credibility because I guess that they have
looked for an expert person or they talk about a book written by a
professional. Internet to me, no. No, no, no. I don’t trust it enough.
20. Results
Patients
Heuristics:
Reputation: recommendations of doctors, patients associations,
friends (personal trustworthiness, cognitive authority), first results
on Google (trust)
Relevance of content (believability)
Journalistic articles (expertise)
Affinity with other patients in web forums (believability)
Checking with self knowledge and self experience with the illness
(believability)
Contrasting with different sources through hyperlinks and search
engines, interactivity... (believability)
21. Results
Patients
Interaction:
•
•
•
expertise clues: medical-research authors, bibliography, scientific
documents
trustworthiness clues: look at the author description, officiality,
recommendations by doctors, friends... delivered by e-mail, social
networks, face to face interaction...
trustworthiness-believability: Homophilia at forums (but weak
credibility), comparisson
22. Results
Expert profile
General profile
•• H1 (context)
H1 (context)
•• H1 (context)
H1 (context)
– Professional conventions
– Professional conventions
•• H2 (verification)
H2 (verification)
– Check with self knowledge
– Check with self knowledge
•• H3 (time)
H3 (time)
– No effect
– No effect
•• H4 (levelling)
H4 (levelling)
– Quick assesment in interaction
– Quick assesment in interaction
– Cognitive authorities
– Cognitive authorities
•• H2 (verification)
H2 (verification)
– Ask doctors
– Ask doctors
– Contrast-hypertext
– Contrast-hypertext
•• H3 (time)
H3 (time)
– Learning and checking
– Learning and checking
•• H4 (levelling)
H4 (levelling)
– General trust in forums
– General trust in forums
– Fear of using information
– Fear of using information
23. Discussion
•
•
•
•
•
Complex and multidimensional concept
Context and user’s skills influence credibility processes
First inventory of heuristics and interaction practices
Caution in consumption
Further research:
– Other informational contexts
– Produsers sample
– Digital natives sample
24. Thank you for your attention!
lluisa.llamero@urv.cat
Blog: http://digicreder.blogspot.com/
Editor's Notes
Hello. I’m here to talk about my thesis project. The object of study is credibility on the Internet and, in particular, today I’m going to talk about a case study that I’ve done in the field of health information.
But before starting I’d like to say that I come from the Universitat Rovira i Virgili of Tarragona. It is a university located in the South of Catalonia-Spain, a one hundred kilometres from Barcelona. The Communication Studies Department works in different areas as risk, corporate, political communication and the Internet. There are 12 doctors and 4 granted PhD students.
My presentation will take around 20 minutes and I’m going to look to these main areas. First, I’ll talk about the importance of the use of internet to gather information about health. After that I’ll explain what is credibility as a concept and why it matters. Then, I’ll briefly talk about different theories of credibility and Internet credibility. Later I’ll describe the methodology and sample of my project. And finally I’ll present the results and discussion about them.
According to Eurostat’ statistics the 38 per cent of the population in the 27 European countries use Internet look for health information, this figure is the average. The countries where this percentage is higher are Iceland, 61, Finland, 58% and Denmark and Germany 54%.
As for Spain, the use is the same of the average, 38%, and in the United Kingdom the rate is 36%.
This numbers show the importance of this resource of information for people nowadays, but let’s take a look at how this way of gathering information is changing our habits in dealing with health information.
Different studies account for a change in the paradigm of health information consumption. The most significant theory tells us that a new kind of patient has arisen: the informed patient. This is a person who is ill, or has relatives or friends with an illness, and he or she looks for information on the Net because is unsatisfied with the information their doctors gave to them. The informed patient use the Internet because is a good resource. Accessibility, free information, access to different sources are some of the advantages. This new practice would be challenging the authority of doctors, or at least patients dare to negotiate with doctors. This can cause some trouble, but on the other hand the positive effect is that being informed helps patients in dealing day by day with their medical condition.
If we look at what doctors do, they are heavy users of the Internet, too. There are few research about this group, but some health institutions have delivered advise about the criteria to assess the information that are based on traditional medical bibliography.
The research in this area and other topic areas, also points out that Internet has arisen important uncertainties to a reasonable consumption like: how to deal with such huge amount of data, how to assess the quality of it because the Web is open and everybody can publish, how to find relevant information in an disorganized system, how to overcome the inadequacy of content to user’s knowledge and what kind of sources deserve trust.
Internet is open and everybody can publish without control or revision. So, the users perceive that the quality of content is highly variable and they must decide what is good information and what is bad information. How they do it? Credibility is a conceptual tool that appears in this process. As a first step, believing that some information could be true helps Internet audiences to filter among the overwhelming volume of data available.
It can act as a tool to assess if the data is good or not for them after they read, listen or watch a document.
Historically, credibility appears as an study object in Aristotle’s works. In our time, the are few studies done by the Mass Communication Research, in the 1950s and after the popularization of Internet a new interest has waken up.
We have several approaches to web credibility. Information literacy, which is more a recommendation to verify a list of criteria than a theory. So, it doesn’t explain what users do to assess credibility. Then, we have the Unifying Framework of Credibility Assessment, the Prominence-Interpretation Theory and the Cognitive Authority. And a series of other studies that explore a different set of variables.
I’ll give some details of the second and the fourth, which are the most useful and have helped me to design my own theoretical framework.
The Unifying Framework of Credibility Assessment is a model proposed by Brian Hilligoss and Soo Young Rieh in 2008.
It thinks in credibility assessment as a multidimensional process with three levels:
first: the construct level, it is the mental base, how the user or a group of users defines credibility according to atributes like truthfulness, expertise or trustworthiness, or others
-second, heuristics, it is a set of practical rules. They come from the construct level or from the experience of the user in consuming web information in an area or topic. For instance, a heuristic to judge reputation is a product of seeing credibility as a question of expertise in the construct level.
Third, the interaction, it is the level where the information consumption takes place and where the users effectively assess the credibility of a content, a media or a particular source.
The CONTEXT is a cross factor that influences the process as a whole: it is not the same consuming information if you are a patient or if you are a doctor, or in other topic areas is not the same reading economic information as a general user than as a professional. The context marks the rules and makes the credibility a SOCIAL process, it is not only a individualistic judgment.
The Cognitive Authority is a social epistemology model proposed by Patrick Wilson in 1983 and it has been used recently by scholars as Rieh, Savolainen (Finland) to put light on web consumption. It has points in common with credibility, but an important nuance, in my view.
According to this theory, information brings knowledge and knowledge can be first or second hand. First hand knowledge is better because is direct. It comes from the things that a person can see and hear without any intermediation. So it is the most reliable for people, but it has a problem: is very limited. If we only knew what we see, we would know few things. So we depend on others to gather information and ideas to give meaning to that information.
Initially there is a suspicion: we don’t believe blindly in all that we hear. The source needs the recognition as an authority to be trusted. But we don’t have enough knowledge so we must to rely on “indirect tests or indexes of credibility”.
According to Wilson a cognitive authority has these characteristics: trustworthiness, truthfulness, scholarship, credibility, oficiality and competence. Now we can see the points in common with the concept of credibility.
The nuance that cognitive authority introduces is a question of grade. A source can be credible and we would say that the information he gives is truth, but if we consider a source as an authority we are going to think that what we gather from that source it is like “self knowledge”.
There are some limitations of the theories and research done up to know. The most important is the inaccuracy in the definition of the concept. There is an ambiguity between believability and credibility. The researchers don’t give accurate or close definition, maybe because it is a concept very wide in meaning.
Another limitation is that there are few studies that address specific Internet credibility assessment and many topic areas are not studied. So we don’t have a model able to put light on different situations of information consumption and is it an important risk to apply the rules of a topic to all the topics.
So, I’ve decided to give an operational definition. An is this:
Although in this presentation I’m only talking about health information, I would like to mention that I’ve gather data from other topic areas as well in my PhD. They are: economic information and travel and leisure information. I’ve collected information from a general and representative sample, but I’ve also done an specific case study of heavy internet users for each topic area.
The approach I’ve chosen is qualitative. It is based on what I call navigational interview. It consists in the traditional semi-structured interview but it’s hold in front of a computer. Why? Because it has two important advantages: for the informants is a help. They can explain what they do, for instance when they look for a query on Google. They can show routines, say in what features they rely on more... And for the researcher this allows to do observation at the same time. So, it combines interviews with think aloud protocol.
And the sample consists of health professionals (family doctors and hospital doctors) and, on the other hand, patients and relatives. I’ve chosen this distinction to test if the context is a factor that influences the criteria performed to assess credibility.
The gathering of data took place from November 2011 to February 2012.
I’m working on four hypotheses: onethe context to try to figure out if professional use or general use produces differences; two,the verification to test if the Internet capabilities produce more actions to verify the information than the tradicional channels; three, the time, is to test if credibility is a question of a specific moment or needs a lot of time to be assessed; and 4 the levelling effect is to figure out if the openness of the Internet causes insecurity and the impression that audiences cannot say for sure if web pages are credible.
Doctors don’t give a single order of attributes which demonstrates that they consider credibility as a multidimensional concept. But it seems that expertise is the most influential attribute. The expertise understood as scientific data and knowledge. This quotation is an example...
Trustworthiness is another attribute essential to assess credibility. It is used to judge the source , the user judge if the source follows a code of ethics, a quality control, etc. We can see the importance in this quotation
This mental structure of the credibility among doctors is translated in several heuristic rules. Most of them are an adaptation of heuristics applied before the Internet. The main heuristics are: the impact factor in medical journals, go to oficial health sites or medical associations, trust in blogs or webs provided by colleagues and examine message signs like the methodology description.
There is a change in grade in two of this heuristics introduce by Internet: the comparison among different sources and the use of Internet’ resources as a reminder for specific information. The immediacy and accessibility of Internet content makes that practice more common.
In the interaction level we have observed the application of most of the heuristics, but we have seen some contradictions too. As a doctor recognised in this quotation: ...
Google is the only search engine used, apart from specific medical databases and many times it brings the same results, to the same medical articles published in journals.
Other uses of Google is as a means of transport to go to trusted sources. It is used because it saves time typing the URLs. In the office it is used in front of the patient to look for fast information without very rigorous verification actions. I have observed it in a query about the characteristics of a drug in one case, and the patterns of a breath problem in another. This practice is based in a confidence in the self knowledge, but it implies risks of credulity and inaccuracy.
If thy don’t have any references of a web, they immediately look for a description of the source.
Patients use the Internet to know more about their medical condition and to look for side questions that they don’t ask doctors in the office because they forgot or they don’t dare to ask. For example how to relax to don’t panic in a colonoscopy or other examinations. They use the Internet to learning about novelties in the treatment of illnesses.
In the construct level they rely on in expertise mainly, so they look for medical an research credentials before reading a document. We can see it in this quotation...
But the other attributes are important too and strongly related.
:
In one case I’ve found a great scepticism on the Internet. The reason is the open nature of the channel. The informant said: ...
The heuristics reported by patients are:
The specific medical information is not evaluated because of the gap in knowledge.
To finish, we have seen that credibility and web credibility is a multidimensional concept with several attributes. That the context and user’s skills influences the process.
We’ve provided a first inventory of heuristics and interaction practices for a specific need of information.
It is important that the general users are cautious in the consumption of information.
And to understand this concept better we must research other information contexts to compare and see if the main attributes are common and if there are heuristics that can be shared.
A gap in the research is in the sample, too. I’ve chosen a representative sample of Internet users randomly, but it would be interesting research heavy users because they could report different heuristics due to their digital skills.
Right, I think that’s everything. And now, if you have any questions, I’ll be happy to try to answer them.