Draft of a talk I helped give to the Summer science communication symposium at Iowa State reporting some initial analyses related to how we should measure trust in the context of public opinion about science and a desire to be more strategic about science communication.
A Strategic Science Communication Approach to Trust
1. Bad/good character Does not have/has integrity Does not/Does let others have a say Insincere/Sincere
Biased/Unbiased Inattentive/Attentive to others Does not/Does keep promises Irresponsible/Responsible
Does not think/
Thinks others' needs are important
Cannot/Can be counted
on to do what's right
Does not/Does keep my/others'
interests in mind
Does not treat/Treats
others with dignity
Does not/Does explain
decisions thoroughly
Not committed/
committed to others
Inexperienced/Experienced Weak/Strong sense of justice
Does not/Does solve problems Cold/Warm
Not conscious/
Conscious of responsibility
Not influenced/
Influence by discussion or feedback
Does not stick/sticks to word Not knowledgeable/Knowledgeable
Dishonest/Honest Not thorough/Thorough
Dishonorable/Honorable Poorly/Well Educated
Disreputable/Reputable Unaccomplished/Accomplished
Disrespectful/Respectful Does not/Does throw weight around Does/Does not withhold information Unclear/Clear
Do/Do not need to keep an eye on Incapable/Capable Hurts/Does not hurt others Unethical/Ethical
Does not/Does tailor
communication to specific audiences
Does not/Does welcome
open discussion or feedback
Does not/Does work
for the good of humanity
Does not/Does take others'
opinions into account
Disloyal/Loyal Does not/Does mislead Impolite/Polite Unintelligent/Intelligent
Does not/Does tell whole story Listens poorly/well Unqualified/Qualified Unprofessional/Professional
Does not/Does put
others interests first
Not concerned/Concerned about
others' interests or well/being
Does not let/
Lets others have influence
Does not/Does believe
are others' views legitimate
Does not care/Cares about others Unreliable/Reliable Immoral/Moral Uninformed/Informed
Inexpert/Expert Does not/Does bend facts Incompetent/Competent Unskilled/Skilled
Does not share/Shares values Does not/Does exaggerate Inconsiderate/Considerate Useless/Useful
Does not/Does improve others' lives Inaccurate/Accurate Inconsistent/Consistent Weak/strong principles
Trust, Credibility, Fairness, and
Deference: Untangling the Variables
Used to Measure Public
Perceptions of Scientists
Nicole Lee, John C. Besley, Geah Pressgrove
2. Bad/good character Does not have/has integrity Does not/Does let others have a say Insincere/Sincere
Biased/Unbiased Inattentive/Attentive to others Does not/Does keep promises Irresponsible/Responsible
Does not think/
Thinks others' needs are important
Cannot/Can be counted
on to do what's right
Does not/Does keep my/others'
interests in mind
Does not treat/Treats
others with dignity
Does not/Does explain
decisions thoroughly
Not committed/
committed to others
Inexperienced/Experienced Weak/Strong sense of justice
Does not/Does solve problems Cold/Warm
Not conscious/
Conscious of responsibility
Not influenced/
Influence by discussion or feedback
Does not stick/sticks to word Not knowledgeable/Knowledgeable
Dishonest/Honest Not thorough/Thorough
Dishonorable/Honorable Poorly/Well Educated
Disreputable/Reputable Unaccomplished/Accomplished
Disrespectful/Respectful Does not/Does throw weight around Does/Does not withhold information Unclear/Clear
Do/Do not need to keep an eye on Incapable/Capable Hurts/Does not hurt others Unethical/Ethical
Does not/Does tailor
communication to specific audiences
Does not/Does welcome
open discussion or feedback
Does not/Does work
for the good of humanity
Does not/Does take others'
opinions into account
Disloyal/Loyal Does not/Does mislead Impolite/Polite Unintelligent/Intelligent
Does not/Does tell whole story Listens poorly/well Unqualified/Qualified Unprofessional/Professional
Does not/Does put
others interests first
Not concerned/Concerned about
others' interests or well/being
Does not let/
Lets others have influence
Does not/Does believe
are others' views legitimate
Does not care/Cares about others Unreliable/Reliable Immoral/Moral Uninformed/Informed
Inexpert/Expert Does not/Does bend facts Incompetent/Competent Unskilled/Skilled
Does not share/Shares values Does not/Does exaggerate Inconsiderate/Considerate Useless/Useful
Does not/Does improve others' lives Inaccurate/Accurate Inconsistent/Consistent Weak/strong principles
A Strategic Science
Communication
Approach to Trust
Nicole Lee, John C. Besley, Geah Pressgrove
4. There is a community of practitioners
desperate for advice about those choices
#Scicomm vs.
#StratSciComm
5. Imagine you’re scheduled for a 60
minute talk (or a 600 word article) …
#Scicomm vs.
#StratSciComm
The science
Risks/BenefitsEfficacy of
Solutions
Researchers'
Motivation
Resaerchers'
Integrity
Researchers'
Competence
Discussion
Also …
• How much time and
resources for preparation?
• Should devote resources to
showing up early/staying late?
• Follow up and Evaluation?
6. Goals should determine communication
objectives; objectives should determine tactics
#Scicomm vs.
#StratSciComm
7. The problem Can I directly communicate
trustworthiness, credibility,
and fairness?
8. The problem (2) It’s important people know I
care. I should probably make
sure I … and tell them about …
9. The research
1. Extract “concept pairs”
from literature
2. Expert sorting
task for concepts
3. Quota-based
online survey
10. Key concepts
(Meyer, P. (1988). Defining and measuring credibility of newspapers. Journalism Quarterly, 65, 567-
574, 588. Used in: McComas, K. A., & Trumbo, C. W. (2001). Source credibility in environmental
health-risk controversies: Application of Meyer's credibility index. Risk Analysis, 21, 467-480.)
14. 1. Bad/good character
2. Biased-Unbiased
3. Cannot/Can be counted on to do
what's right
4. Cold-Warm
5. Dishonest/Honest
6. Dishonorable-Honorable
7. Disloyal-Loyal
8. Disrespectful-Respectful
9. Do not/Have confidence in ...
10.Do/Do not need to keep an eye on
11.Does not care/Cares about others
12.Does not have/has integrity
13.Does not let/Lets others
have influence
14.Does not meet/
Meets people's needs
15.Does not share/Shares values
16.Does not stick/sticks to word
17.Does not think/Thinks others'
needs are important
18.Does not treat/Treats
others with dignity
19.Does not try/tries to
avoid hurting others
20.Does not try/tries to
improve others’ lives
21.Does not/Allow creativity
22.Does not/Does believe others'
views legitimate
23.Does not/Does bend facts
24.Does not/Does communicate
in a timely way
25.Does not/Does exaggerate
26.Does not/Does explain
decisions thoroughly
27.Does not/Does keep my/others'
interests in mind
28.Does not/Does keep promises
29.Does not/Does
let others have a say
30.Does not/Does meet expectations
31.Does not/Does put
others interests first
32.Does not/Does sensationalize
33.Does not/Does sensationalize
34.Does not/Does solve problems
35.Does not/Does tailor
communication to specific
audiences
36.Does not/Does take others'
opinions into account
37.Does not/Does tell whole story
38.Does not/Does throw
weight around
39.Does not/Does welcome open
discussion or feedback
40.Does not/Does work
for the good of humanity
41.Does/Does not withhold
information
42.Immoral/Moral
43.Impolite-Polite
44.Inaccurate-Accurate
45.Inattentive/Attentive to others
46.Incompetent-Competent
47.Inconsiderate-Considerate
48.Inconsistent-Consistent
49.Inexperienced-Experienced
50.Inexpert-Expert
51.Insincere-Sincere
52.Intolerant-Tolerant
53.Irresponsible-Responsible
54.Listens poorly/well
55.Misleads/Does not mislead
56.Not candid-Candid
57.Not capable-Capable
58.Not committed/
committed to others
59.Not concerned/Concerned about
others' interests or well-being
60.Not conscious/Conscious of
responsibility
61.Not influenced/Influenced by
discussion or feedback
62.Not knowledgeable-Knowledgeable
63.Not reputable-Reputable
64.Not thorough-Thorough
65.Poorly/Well Educated
66.Should not/should influence
decisions
67.Unaccomplished-Accomplished
68.Unbelievable-Believable
69.Unclear/Clear
70.Unconvincing-Convincing
71.Uncooperative-Cooperative
72.Unethical-Ethical
73.Unfair-Fair
74.Unhelpful/Helpful
75.Uninformed-Informed
76.Unintelligent-Intelligent
77.Unprofessional/Professional
78.Unqualified-Qualified
79.Unreasonable-Reasonable
80.Unreliable-Reliable
81.Unskilled-Skilled
82.Untrustworthy-Trustworthy
83.Useless-Useful
84.Weak/strong principles
85.Weak/Strong sense of justice
15. 1. Bad/good character
2. Biased-Unbiased
3. Cannot/Can be counted on to do
what's right
4. Cold-Warm
5. Dishonest/Honest
6. Dishonorable-Honorable
7. Disloyal-Loyal
8. Disrespectful-Respectful
Do not/Have confidence in ...
Do/Do not need to keep an eye on
9. Does not care/Cares about others
10.Does not have/has integrity
11.Does not let/Lets others
have influence
Does not meet/
Meets people's needs
12.Does not share/Shares values
Does not stick/sticks to word
13.Does not think/Thinks others'
needs are important
14.Does not treat/Treats
others with dignity
15.Does not try/tries to
avoid hurting others
16.Does not try/tries to
improve others’ lives
Does not/Allow creativity
17.Does not/Does believe others'
views legitimate
Does not/Does bend facts
Does not/Does communicate
in a timely way
Does not/Does exaggerate
18.Does not/Does explain
decisions thoroughly
19.Does not/Does keep my/others'
interests in mind
20.Does not/Does keep promises
21.Does not/Does
let others have a say
Does not/Does meet expectations
Does not/Does put
others interests first
Does not/Does sensationalize
Does not/Does sensationalize
Does not/Does solve problems
22.Does not/Does tailor
communication to specific
audiences
23.Does not/Does take others'
opinions into account
24.Does not/Does tell whole story
Does not/Does throw
weight around
25.Does not/Does welcome open
discussion or feedback
26.Does not/Does work
for the good of humanity
27.Does/Does not withhold
information
28.Immoral/Moral
29.Impolite-Polite
30.Inaccurate-Accurate
31.Inattentive/Attentive to others
32.Incompetent-Competent
33.Inconsiderate-Considerate
Inconsistent-Consistent
34.Inexperienced-Experienced
35.Inexpert-Expert
36.Insincere-Sincere
Intolerant-Tolerant
37.Irresponsible-Responsible
38.Listens poorly/well
39.Misleads/Does not mislead
Not candid-Candid
40.Not capable-Capable
Not committed/
committed to others
41.Not concerned/Concerned about
others' interests or well-being
Not conscious/Conscious of
responsibility
Not influenced/Influenced by
discussion or feedback
42.Not knowledgeable-Knowledgeable
Not reputable-Reputable
43.Not thorough-Thorough
Poorly/Well Educated
44.Should not/should influence
decisions
Unaccomplished-Accomplished
45.Unbelievable-Believable
46.Unclear/Clear
47.Unconvincing-Convincing
Uncooperative-Cooperative
48.Unethical-Ethical
49.Unfair-Fair
Unhelpful/Helpful
50.Uninformed-Informed
51.Unintelligent-Intelligent
52.Unprofessional/Professional
53.Unqualified-Qualified
Unreasonable-Reasonable
54.Unreliable-Reliable
55.Unskilled-Skilled
56.Untrustworthy-Trustworthy
57.Useless-Useful
Weak/strong principles
58.Weak/Strong sense of justice
Culled by logic,
obvious duplication
16. Asking the experts – What did it tell us?
Scholars with related expertise (N = 7) grouped items into categories
• Patterns emerged but no
consensus
• Helped cull ambiguous items
• Most came up with 4 or 5
sub-categories
• Too many competence/
expertise items
17. Survey work
• Qualtrics quota sample (N = 600)
• Randomly assigned to a generic
scientist or GMO-scientist version
• 58 trust questions randomly ordered
• Additional blocks on …
• attitudes about science
• specific scientific issues
• science knowledge
• CFAs (MLR) on general scientist sub-
sample based on sorting and theory
• Test 2-, 3-, and 4-factor models
• Pare items
• Replication using GMO sub-sample
18. Competence/Ability Standardized
Estimate
Integrity/Character Standardized
Estimate
Incompetent-Competent .77 Dishonest-Honest .80
Incapable-Capable .79 Dishonorable-Honorable .80
Unskilled-Skilled .83 Bad-Good Character .74
Not-Knowledgeable .81 Does not have-Has integrity .77
Inexperienced-Experienced .76 Does not-Does … values .68
Inexpert-Expert .79 Does not-Does … promises .73
Unprofessional-Professional .80 Does not-Does … whole story .70
Unintelligent-Intelligent .69 Immoral-Moral .81
Not-Thorough .78 Insincere-Sincere .83
Uninformed-Informed .76 Does not-Does … mislead .59
Unqualified-Qualified .82 Unethical-Ethical .81
RMSEA (90 CI)
Probability of P < .05
.04 (.02-.06)
.70
.05 (.02-.07)
.50
CFI .98 .98
SRMR .03 .03
Individual
measures
General science sub-sample, n = 293
(Red denotes cut with major modifications)+ Replicated with GMO sub-sample
19. Benevolence/Warmth Standardized
Estimate
Respectful/Listens Standardized
Estimate
Does not care-Cares .82 Poorly-Well … listen .77
Does not think-Thinks … needs .74 Does not let-Lets … have say .71
Hurts-Does not hurt … others .53 Does not see-See … legitimate .74
Does not improve-Improves .76 Disrespectful-Respectful .78
Does not keep-Keeps … interests .74 Does not let-Lets … influence .70
Not working-Working … humanity .70 Does not-Does welcome .89
Unconcerned-Concerned … others .74 Does not take-Takes … views .81
Unconcerned-Concerned … well being .84 Impolite-Polite .76
Inattention-Attention … others .71
Does not tailor-Tailors .64
RMSEA (90 CI)
Probability of P < .05
.00 (.00-.03)
.99
.06 (.03-.08)
.32
CFI 1.00 .97
SRMR .02 .04
Individual
measures
General science sub-sample, n = 293
21. Modeling: Attitudes toward science in general
(five items, alpha = .74, Range 1-5, adjusted r2 increase from trust items = .28)
-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
Two Factor Model
Three Factor Model
Four Factor Model
Science Knowledge
Perceived Competence
Perceived Listening
or All Character Variables
Perceived Integrity or
Integrity with Benevolence
Perceived Benevolence
Controlling for condition, education, Hispanic, White, ideology, and religiosity
*
*
*
*
22. Modeling: Oppose/Support Nuclear Energy
(Single item, Range 1-5, adjusted r2 increase from trust items = .01)
Science Knowledge
Perceived Competence
Perceived Listening
or All Character Variables
Perceived Integrity or
Integrity with Benevolence
Perceived Benevolence
Controlling for condition, education, Hispanic, White, ideology, and religiosity
*
*
*
*
-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
Two Factor Model
Three Factor Model
Four Factor Model
Attitudes towards Science
23. Modeling: Oppose/Support Genetically Engineered Food
(Single item, Range 1-5, adjusted r2 increase from trust items = .04)
Science Knowledge
Perceived Competence
Perceived Listening
or All Character Variables
Perceived Integrity or
Integrity with Benevolence
Perceived Benevolence
Controlling for condition, education, Hispanic, White, ideology, and religiosity
*
Attitudes towards Science
-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
Two Factor Model
Three Factor Model
Four Factor Model
*
*
*
*
*
*
*
*
*
24. Modeling: Oppose/Support Use of Autonomous Vehicles
(Single item, Range 1-5, adjusted r2 increase from trust items = .03)
Science Knowledge
Perceived Competence
Perceived Listening
or All Character Variables
Perceived Integrity or
Integrity with Benevolence
Perceived Benevolence
Controlling for condition, education, Hispanic, White, ideology, and religiosity
*
*
*
*
-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
Two Factor Model
Three Factor Model
Four Factor Model
Attitudes towards Science
*
*
*
*
*
25. Modeling: Oppose/Support Climate Change Action
(Single item, Range 1-5, adjusted r2 increase from trust items = .01)
Science Knowledge
Perceived Competence
Perceived Listening
or All Character Variables
Perceived Integrity or
Integrity with Benevolence
Perceived Benevolence
Controlling for condition, education, Hispanic, White, ideology, and religiosity
Attitudes towards Science
-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
Two Factor Model
Three Factor Model
Four Factor Model
*
*
*
*
Trust, Credibility, Fairness, and Deference: Untangling the Variables Used to Measure Public Perceptions of Scientists
Trust, Credibility, Fairness, and Deference: Untangling the Variables Used to Measure Public Perceptions of Scientists
People have much more positive views about their audiences in the context of F2F engagement.
Key points
Goals … are audience specific
Theory … tell us what objectives are most likely to result in goals
Key points
Goals … are audience specific
Theory … tell us what objectives are most likely to result in goals
One key set of objectives that can help us accomplish a lot of goals are related to trust
For any given objective, there are lots of potential tactics/variation on tactics
Key points
Goals … are audience specific
Theory … tell us what objectives are most likely to result in goals
One key set of objectives that can help us accomplish a lot of goals are related to trust
For any given objective, there are lots of potential tactics/variation on tactics
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
Too many competence questions …
Planned and still hope to recruit more experts (like all of you)
Only analyzed qualitatively but need to figure out formal analysis for larger sample
Helped inform measurement models
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
What would a message that communicates trust (and trust alone) look like …
4 variables excluded for conceptual reasons (trust, fairness, bias), 14 – variables didn’t work as part of single-measure CFA
4 factor modified – correlations – listen/character = .87, listen/benevolence = .89, character/benevolence = .93, listen/able = .69, character/able = .86, benevolence/able = .82.
3 factor modified – correlations – listen/character = .91, listen/able = .72, character/able = .87.
2 factor modified – correlation = .82
- I can make a supercut where I get a 2-factor model down to r = . 84.
4 Factors – Modifications
(xx variables cut)
191.11
146
1.31
12429.00
12461.06
.03 (.02-.05)
.98
.03
Lesson: Non-competence trust only reveals itself in the two-factor model
Lessons: Need to combine the non-competence trust measures to reach significance.
Lessons: Integrity is the thing for GMO (but lumps together with other character traits; makes competence negative (you have to believe scientists can make tricky things …)
Lesson: Non-competence trust only reveals itself in the two-factor model
Lesson: Perceived willingness to listen might matter on its own but is washed out by other variables ???