|
|
|
|
Practical Information
|
Instrument Name:
|
Organizational Trust Inventory (OTI)
|
Instrument Description:
|
This 81-item instrument was designed to measure the degree of trust between units of an organization or between organizations. The authors define trust as an individual’s or group’s belief that another individual or group "makes good-faith efforts" to honor commitments, is “honest in negotiations that preceded those commitments”, and “does not take excessive advantage of another”, at the expense of another, when such an occasion to do so exists. These three dimensions along with three components of assessment (affective, cognitive, behavioral) constitute a matrix in which the items fall. (Ref: 1)
|
Price:
|
Free for scholarly uses; negotiable for for-profit uses
|
Administration Time:
|
Approximately 25 minutes (long form) or 5 minutes (short form)
|
Publication Year:
|
1995
|
Item Readability:
|
Flesch-Kincaid Grade Level of 6.6. Items have simple sentence structure. Some vocabulary is organization-related (e.g. "negotiations," "obligations").
|
Scale Format:
|
7-point Likert-type scale; some questions are negatively worded.
|
Administration Technique:
|
Self-administered; no specific requirements
|
Scoring and Interpretation:
|
Items are added. Higher numbers indicate higher levels of trust. Some items have reverse scoring.
|
Forms:
|
A 12-item short form of the OTI is also available.
|
|
|
Research Contacts
|
Instrument Developers:
|
L.L. Cummings and Philip Bromiley
|
Instrument Development Location:
|
(Bromiley)
C L Carlson School of Management
321 19th Ave S, 3-365
Minneapolis, MN 55455
|
Instrument Developer Email:
|
pbromiley@csom.umn.edu
|
Instrument Developer Website:
|
No information found.
|
|
|
Annotated Bibliography
|
1. Cummings LL, Bromiley P. The Organizational Trust Inventory (OTI). Trust in organizations: Frontiers of theory and research. Kramer RM, Tyler TR (eds.) Sage Publications, Inc. 1996: 302-330.
 |
Purpose: The authors intended to design an instrument to measure organizational trust. |
|
Sample: Responses from 323 employees and students from undergraduate, MBA, and executive MBA programs from the University of Minnesota were used. |
|
Methods: 273 items related to the three dimensions and following criteria set by the authors were developed by five doctoral students over a six-month period. These were reduced to 121, combined with 15 items from a previously validated Organizational Commitment Questionnaire (OCQ; a related construct) and administered to three different doctoral students for categorization into the three dimensions. The results showed good dimensional differentiation without confusion with the OCQ’s dimension. Similarly worded items were eliminated equally and a final count of 62 trust questions and 19 behavioral questions. |
|
Implications: Previously developed trust construct measures were not validated. |
|
|
top
|
Factors and Norms
|
Factor Analysis Work:
|
The authors performed a confirmatory factor analysis using latent variable structure equation analysis from the 323 responses. Estimation analysis was performed in three stages: estimation of items versus the three dimensions of trust; estimation for each dimension based on response-mode factors (i.e. affect, cognitive and intent); and estimation of the model with both the trust factors and the reported behavior. (Ref: 1)
|
Normative Information Availability:
|
No information found.
|
|
|
Reliability Evidence
|
Test-retest:
|
No information found.
|
Inter-rater:
|
During the developmental stage (using 136 items) the authors asked three doctoral students to classify the items into one of four categories each. They found that all three sorters agreed unanimously on 85.3% of the items, and two of three agreed on 14.7% of the items. There were no complete disagreements. All Cohen’s kappa values were above 0.83. (Ref: 1)
|
Internal Consistency:
|
Composite reliability was calculated for the initial estimation with all questions and then separately for each of the three dimensions with the response-mode factors (RMF.) Estimation result reliabilities were above 0.94. The other reliabilities were as follows (Ref: 1):
|
|
Keeps Commitments
|
Negotiates Honestly
|
Avoids Taking Advantage
|
| Affect
|
0.89
|
0.92
|
0.89
|
| Cognitive
|
0.49
|
0.94
|
0.92
|
| Intent to Behave
|
0.40
|
0.77
|
0.88
|
|
Alternate Forms:
|
No information found.
|
|
|
Validity Evidence
|
Construct/ Convergent/ Discriminant:
|
The authors evaluated whether the OTI items could be distinguished from items of a related construct. The 15 items from the Organizational Commitment Questionnaire were included with 121 items from the OTI and were sorted into categories by raters. The available categories were the 3 trust dimensions (as described in the Instrument Description section) and a “commitment category”. None of the OTI items were categorized as commitment items by any of the raters. (Ref: 1)
|
Criterion-related/ Concurrent/ Predictive:
|
Items were constructed for self-reported behaviors and these items were included on the questionnaire with the OTI items. The Associations between these behavior scores and the trust scores were high, however, the behavior items are very similar to the OTI items and such correlations are not unexpected. For example, one of the OTI items is, “We feel we can depend on ____ to move our joint projects forward.” One of the behavior items is, “We check to make sure that _____ continues to work on our joint projects.”
|
Content:
|
The use of a definitional matrix in developing items supports the content validity of the measure.
|
Responsiveness Evidence:
|
No information found.
|
Scale Application in VA Populations:
|
No information found.
|
Scale Application in non-VA Populations:
|
Yes. (Ref: 1)
|
|
|
Comments
|
In developing the OTI, the authors were methodical in defining the construct, “organizational trust” and situate their definition within the context of organizational theory. This is particularly important for the measurement of a psychosocial variable such as organizational trust for which definition is no trivial task. Another strength of the measure is that the items were developed in a systematic way, using what the scale developers call a “definitional matrix of trust as belief”. This matrix functions as a test blueprint and ensures that the domain being measured is well-represented by the items of the OTI. The developers calculated the associations between a set of ad hoc “behavior measures” and OTI scores. Because the content of the behavior items and the OTI items was so similar, the results are unsurprising and offer only weak evidence for the concurrent validity of the scores.
The distinction between affective and cognitive modes of response was not supported by the initial results, and some items had low loadings on their respective factor. The difficulty of measuring a construct such as organizational trust should be considered in evaluating these results.
|
Updates
No information found.
|
|
|
