TrustRisk Raw data file column key

The following are explanations of the standard column labels used in all of the Trust & Risk raw data .txt files. Where any given experiment includes additional raw data, the extra columns are explained in the summary file in that experiment's folder.


Number:

Number assigned to each participant, after they have been organized into groups.


Code:

The code, made up of 3 or 4 letters and 3 numbers, which was used to assign participants to experimental groups. Each letter position is referred to in the experimental summary file as "codes" - i.e. the first letter is code1, second letter is code2, and so on. Each unique letter combination represents an experimental group, so participants with the same letters are in the same group.


Friend:

In experiments 6-10, 16 & 18-19 a name is listed in this column. This is not the participant’s name or indeed any real name, but that of an imaginary or hypothetical friend to whom they might give advice about risk.



Trial:

Identifies judgment trials in the (randomized) order that they were run.


Sex:

Gender of participant, where 1 = female, and 2 = male.


Age:

Age of participant.


Criterion:

This is the mean historical value for any given risk type. Specific government and consumer estimates were drawn from normal distributions based around the criterion.


Gov:

Government advisor estimate for any given  risk activity, drawn from a normal distribution around the criterion value one column to the left.


Con:

Consumer advisor estimate for any given  risk activity, drawn from a normal distribution around the criterion value two columns to the left.


Judgment:

Risk estimate or other judgment value entered by the participants for any given trial.


Denominator:

The number by which participants’ risk judgments are divided in order to arrive at the actual risk estimation.



Item:

A number which identifies the item being responded to in any given trial. This identifier is required as the item order was randomized in all experiments, therefore 'Trial no.' and 'Item ID' will almost certainly be different.


Domain:

Refers to the domain of activity that applies to each row of explicit judgments to the right of this column. The first group is judgments on government advisors, the second group is judgments on consumer advisors. The first row in each group is judgments on that generic category of advisor, and subsequent rows within each group are judgments on specific advisors seen during the course of the experiment. This column may appear to the right of the explicit judgments (Trust etc) in some files.

Trust:

An explicit measure of the participants' trust in the advisor, rated on a scale from 1 to 7, where 1 denotes complete trust, and 7 denotes a complete lack of trust.


Bias:

An explicit measure of perceived advisor bias, on a range from -50 (i.e. advisor systematically under-estimates risk by 50%) to +50 (i.e. advisor systematically over-estimates risk by 50%).


Accuracy:

An explicit measure of perceived advisor accuracy, on the same 7-point scale as 'trust'; i.e. lower numbers denote a greater degree of advisor accuracy.


Values:

The first of the six 'Similarity of Values' index items. An explicit measure of perceived similarity between an advisor's values and those of the participant, on the same 7-point scale as 'trust'; i.e. lower numbers denote a greater degree of perceived similarity.


Line (or “in line with me”):

A Similarity of Values index item on a seven point scale, where lower numbers denote a greater tendency for advisors to be "in line" with the participant.


Goals:

As above, but referring to the similarity of advisor & participant goals.


Views:

As above, but referring to the similarity of advisor & participant view.


Acts:

As above, but referring to the similarity of advisor & participant actions.


Thinks:

As above, but referring to the similarity of advisor & participant thinking.






NOTE:  Experiment 11 also includes a column named “bet”, which records the amount participants’ would bet on the outcome of a particular football match. This experiment also records the Yes (Y) & No (N) answers given in response to each of three matches (in the appropriate condition) in columns marked “match1”, “match2”, & “match3”.



NOTE:  Experiments 17 & 21-22 include 4-12 columns (depending on how any given spreadsheet program interprets spaces) set aside for comments on the physical & value similarity settings of the two advisors. These comments are on the right-hand edge of the file, aside from all other data.