FoQus for TRaC Improvement
PSI launched its initiative to start using multi-item scales in the summer of 2004. Multi-item scales are used in TRaC surveys and are useful for measuring concepts like attitudes and beliefs that cannot be directly observed.
FoQus for TRaC Improvement improves the process for multi-item scale development and adaptation within the PSI Behavior Change Framework. Designed to increase the level of input from members of the target audience, FoQus for TRaC Improvement uses qualitative data collection methods to improve the reliability and validity of scales.
Multi-Item Scales
About Multi-Item Scales
Multi-item scales consist of a number of statements that allow for the measurement of complex concepts that are not easily measured by a single question. These statements can be used to measure variables that are not directly observable (they can not be seen) and may change over time. So, while it is possible to directly observe some variables, such as behavior or condom availability in a store or hotspot, a theoretical variable, such as attitudes towards condom use, cannot be directly observed. In TraC surveys, these theoretical variables are determinants of behavior. Examples of determinants that can be measured by multi-item scales include attitudes towards condom use with trusted partners, social norms towards using mosquito nets in a community, self-efficacy for condom use, or social support for family planning.
Defining Characteristics
Multi-item scales consist of a related set of statements which measure a single variable or determinant. There are at least three defining characteristics of multi-item scales (Spector, 1992).
1. The scale is comprised of a minimum of three statements. Less than three statements would be analyzed as individual statements
2. The scale measures a variable/determinant that is continuous and responses can range from very positive to very negative.
3. The scale provides response options designed to capture a broad spectrum of opinions, attitudes, and/or beliefs.
Spector, P. E. (1992). Summated rating scale construction: An introduction. Newbury Park, CA: Sage Publications.
Example
Example of multi-item scales measuring the determinant attitude.
Otto-Salaj, L. L., Kelly, J. A., Stevenson, L. Y., Hoffmann, R., & Kalichman, S. C. (2001). Outcomes of a randomized small-group HIV prevention intervention trial for people with serious mental illness. Community Mental Health Journal, 27(2), 123-144.
*Scale modified for use by PSI.
Qualitative Data Collection Tools
FoQus for TRaC Improvement uses one-on-one interviews as well as targeted activities such as free-listing, pile-sorts, and focus group discussions to gather information for scale modification and development. Group or cognitive interviews are used to pre-test scales.