Welcome to the Social and Behavioral Instruments (SABI) database provided to you by the UNC Center for AIDS Research Social and Behavioral Sciences Research Core! We created this database to assist you in finding the measurement instruments to answer your research questions related to HIV prevention research. We are appreciative of the assistance that the Duke CFAR Social and Behavioral Sciences Core provided in identifying measures for the database. Please contact us at firstname.lastname@example.org if you have questions or need any assistance in searching through our database.
Also, please remember to cite the UNC CFAR: P30-AI50410.
About SABI Database
The idea behind the SABI Database is to provide a user-friendly database of instruments (also called measures or scales) measuring social and behavioral constructs relevant to HIV research that researchers can use to find and compare measures. The database presents a comprehensive list of instruments which have been used to measure certain domains in HIV research.
The SBSRC team used Google Scholar to search for scales using the names of the domains (please see "SABI domains" on the SABI home page) and the key word "HIV". We have included the most relevant source article for each measure. Please see the "Tips for Selecting Measures" tab for more information on how to find other articles describing particular measures in populations and settings that are of interest to you.The SABI contains information on the following characteristics of each measure:
- Psychometric properties (reliability and validity), if available
- Year of publication
- Number of items in the measure
- Population the measure was tested with
- Languages the measure supports
- Citation for the original article that lists the measure items
Researchers may use their own institutional permissions to access the measures listed in SABI by searching for the corresponding articles in Google Scholar or PubMed.
If you would like to recommend that a certain measure, article, or domain be added to the SABI database, please email your recommendation to Catherine Grodensky, Core Manager.
Please remember to cite the UNC CFAR: P30-AI50410
The database has twelve domains. We use the term "domain" to refer to the HIV-related psychosocial processes, conditions, behaviors, and attributes we have selected to include in the SABI database. Please click here to view the SABI domains.
The database has twelve domains. hide
- Social Support
- Quality of Life
- AIDS-related Stigma
- Self-efficacy to disclose HIV status
- Medication adherence self-efficacy
- Safer sex self-efficacy
- Condom use attitudes
- Treatment and transmission risk perceptions
- HIV testing attitudes
- Mental Illness
- Bipolar Disorder
- Anxiety disorders
- Post Traumatic Stress Disorder (PTSD)
- Substance Use
- Adherence to HIV Antiretroviral Therapy (Medication Adherence)
- Sexual Risk Behavior
- HIV Knowledge
- Interpersonal Violence
Definitions for the domains in the database view
- Social Support
The individual belief that one is cared for and loved, esteemed and valued, and belongs to a network of communication and mutual obligations
Cobb S. Social support as moderator of life stress. Psychosomatic Medicine, 1976; 38: 300-314.
- Quality of Life
Individuals' perceptions of their position in life in the context of the culture and value system in which they live, and in relation to their goals, expectations, standards and concerns. It is a broad ranging concept, incorporating in a complex way a person’s physical health, psychological state, level of independence, social relationships, personal beliefs and relationship to salient features of the environment.
Reference: Quality of Life Assessment. The WHOQOL Group, 1994. What Quality of Life? The WHOQOL Group. In: World Health Forum. WHO, Geneva, 1996.
- AIDS-related Stigma
AIDS-related stigma and discrimination refers to prejudice, negative attitudes, abuse and maltreatment directed at people living with HIV and AIDS. These processes can result in the stigmatized person being shunned by family, peers and the wider community; poor treatment in healthcare and education settings; an erosion of rights; and psychological damage; and can negatively affect the success of testing and treatment.
Perceived self-efficacy is defined as people's beliefs about their capabilities to produce designated levels of performance that exercise influence over events that affect their lives. Self-efficacy beliefs determine how people feel, think, motivate themselves and behave. Such beliefs produce these diverse effects through four major processes. They include cognitive, motivational, affective and selection processes.
Bandura, A. (1994). Self-efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of human behavior (Vol. 4, pp. 71-81). New York: Academic Press. (Reprinted in H. Friedman [Ed.], Encyclopedia of mental health. San Diego: Academic Press, 1998.
An attitude is a hypothetical construct that represents an individual's degree of like or dislike for an item. Attitudes are generally positive or negative views of a person, place, thing, or event-- this is often referred to as the attitude object. People can also be conflicted or ambivalent toward an object, meaning that they simultaneously possess both positive and negative attitudes toward the item in question.
The psychological definition of coping is the process of managing taxing circumstances, expending effort to solve personal and interpersonal problems, and seeking to master, minimize, reduce or tolerate stress or conflict.
-Weiten, W., & Lloyd, M. A. (2006) Psychology Applied to Modern Life. Thomson Wadsworth; Belmont California.
-Snyder, C. R. (editor) (1999) Coping: The Psychology of What Works. New York: Oxford University Press.
-Zeidner, M. & Endler, N. S. (editors) (1996) Handbook of Coping: Theory, Research, Applications. New York: John Wiley.
- Mental Illness
How a person thinks, feels, and acts when faced with life's situations. Mental health is how people look at themselves, their lives, and the other people in their lives; evaluate their challenges and problems; and explore choices. This includes handling stress, relating to other people, and making decisions.
Mental Health Dictionary provided by SAMHSA Health Information Network http://mentalhealth.samhsa.gov/resources/dictionary.aspx#M
Mental Health domain in this database includes instruments to measure bipolar disorder, anxiety disorders, posttraumatic stress disorder (PTSD) and depression.
- Substance Use
- Use and misuse of alcohol and other drugs.
- Adherence to HIV Antiretroviral Therapy
Medication adherence may be defined as the extent to which a patient takes a medication in the way intended by a health care provider.
HIV InSite Knowledge Base Chapter May 2005; Content reviewed January 2006
Edward L. Machtinger, MD, University of California San Francisco
David R. Bangsberg, MD, University of California San Francisco
- Sexual Risk Behavior
- Assessment of sexual behaviors which may lead to acquiring or transmitting HIV
- HIV Knowledge
- Knowledge about HIV/AIDS including methods of transmission and prevention.
- Interpersonal Violence
- Experience or perpetration of behaviors that cause physical, psychological, or sexual harm.
How to access
Anyone can create an account, but only users affiliated with UNC CFAR will have full access to the SABI database. Full access means the ability to request consultation and measures directly from the UNC CFAR Social and Behavioral Science Research Core.UNC CFAR affiliated institutions are:
- University of North Carolina at Chapel Hill
- Family Health International
- RTI International
- Other CFARs
- North Carolina-based Historically Black Colleges and Universities
If your institution is affiliated with UNC CFAR and you have difficulties with access to the database please contact us at email@example.com
Tips for Selecting a Measure
Please consult with the CFAR Social and Behavioral Core at firstname.lastname@example.org if you need any support during the instrument selection process.
It is critical to evaluate how well a measure fits the actual variable you intend to examine. The measure you use will determine the variable you will actually quantify. The data that you collect can only reflect the true relationships between your variables of interest if your measures accurately correspond to these variables. A poor choice in measures, therefore, can severely undercut your capacity to study the variables of concern to you.
We recommend considering the following seven steps when selecting an instrument:
- Think about what variable(s) you want to measure.view
You may have already identified your variables of interest by name (e.g., HIV related stigma), but it's crucial to have a more detailed understanding of precisely what it is you want to measure. It is very useful to write a short paragraph defining the variable in which you are interested, how it relates to other variables, and what aspects of it are critical or central to your interests. This will help you to avoid selecting a measure with a name that sounds relevant but with content that does not match your research interests.
- Review the SABI domains to determine whether they include your variables of interest.view
The SABI database has twelve domains. We use the term
domainto refer to the categories of HIV-related psychosocial processes, conditions, behaviors, and attributes we have selected to include in the SABI database. A list of the SABI domains and their definitions is available on the home page of the SABI website under the SABI Domains tab. Reading the domain descriptions is a useful early step in determining whether this site can direct you to the measure that you require. For example, if you are interested in looking for different ways to measure viral load, you would see that we do not provide information about clinical lab measures by looking at the domains. However, if you are interested in some aspect of attitudes towards HIV testing, you would see the domain
attitudesand recognize that examining the measures in that domain would probably be worthwhile.
Some of the SABI database domains such as
mental illness, have sub-domains. If you are interested in measuring a construct from one of these domains, you may want to specify more detail related to the construct. For example, if you want to measure self-efficacy, you will first need to think about what behavior or behavioral category you are interested in, and then specify self-efficacy to perform this particular behavior or category (e.g.
self-efficacy to practice safer sex,
self-efficacy to adhere to medications,etc).
- Conduct a
Search by Keywordor
Advanced Searchto find potential measures of interest.view
It is usually better to keep your search more general to avoid missing measures that may be of interest to you.
Conduct a Search by Keyword if you know the exact name of the construct you want to measure, the name of the author of the measure, or the name of one specific measure you are looking for. Type this information into the "Search by Keyword" space in the SABI Search page. Conduct an Advanced Search if you want to select from several specific search criteria such as domains, psychometric properties, or measures used for specific populations. This search option can be useful if you are still thinking about what you want to measure. To conduct an advanced search, click the
Advanced Searchlink on the SABI Search page and specify the search criteria of interest to you.
When you are finished specifying your search criteria, click submit. Review the results of your search on the SABI Results screen. You can sort the results by clicking any of the column headers. More results may be available than the ones displayed on the page - check the top right corner of the search results page to change the number of items displayed or to view additional search results. The description of each scale is truncated for ease of browsing the results - to view the entire contents of a field, click more.
- Consider the scale's track record for reliability and validity under circumstances similar to your intended use of the instrument.view
It is very important to remember that there is no such thing as a reliable or a valid scale. Reliability and validity are not properties of measures that apply across contexts. They are properties of the measures on specific occasions under particular circumstances. If an instrument has demonstrated good reliability and validity under circumstances that are very similar to those in which you plan to use it, then that instrument will probably work well for you. On the other hand, if the circumstances in which you will use a scale are very different from those in which its reliability and validity have been assessed in the past, then it may not perform similarly. Ultimately, it is your responsibility to determine whether a scale is reliable and valid in the situation in which you are using it.
Reliability is the extent to which variability in the scores an instrument yields is due to actual variability in the construct that the instrument measures. In other words, it represents the absence of random error in the instrument's scores. There are two common ways to estimate reliability: Test/Retest and Internal Consistency.Test/Retest
To estimate test/retest reliability, the same test is administered to the same individual at two separate times. Then, correlation is calculated between the measurements for each individual subject. A critical assumption of test-retest reliability is that the variable being assessed has not changed over the time between assessments. A shortcoming of test-retest reliability is that it confounds changes in scores that arise from random error with changes that occur from real changes in the variable. Only when the assumption of "no change in the variable" holds, will a test-retest correlation be indicative of the reliability of the instrumentInternal Consistency
Internal consistency is often measured through a Cronbach's Alpha score. This statistic estimates the degree to which all items of a scale measure
the same thing. A critical assumption underlying Cronbach's alpha is that the items making up the scale are unidimensional; that is, that they all share one and only one underlying variable. If this is not the case (e.g., if there are two discernable dimensions, factors, or variables underlying the items comprising a scale), the score you will get for Cronbach's alpha will be inaccurate and probably inflated. If an instrument is a multidimensional scale (for example, measuring the strength of health provider characteristics with some questions and the importance of those characteristics with other questions), then each homogeneous group of items is really a separate subscale and Cronbach's alpha should be computed for each of the subscales separately. By convention, a Cronbach’s alpha score is generally expected to be equal to or greater than .70 to indicate adequate reliability, although how much reliability is sufficient for a scale will depend on the context in which it will be used.
Note that some degree of redundancy is likely necessary in a set of items that are all about the same thing. For example, a scale about emotional well-being might include two items reading
I felt down in the dumpsand
I felt blue. However, while concept redundancy is necessary, redundancy in ways unrelated to the construct (such as grammatical structure, use of common incidental words or phrases across items) probably is unnecessary. In general, more items will increase measurement precision in the same way that a larger sample will provide a more precise estimate of the population you are studying.
Validity concerns whether a measure captures the phenomenon it is intended to assess. It is generally classified into three broad types: content, criterion and construct validity.
Content validity concerns how well the selected scale items cover all the dimensions of the relevant construct (i.e., the extent to which the content of the item set corresponds to the definition of the variable under study). Claims of content validity are often supported by citation of relevant literature and/or review by a panel of experts.
Criterion validity concerns the extent to which scores on an instrument compare to a reference standard (i.e., criterion). Criterion validity is largely atheoretical. It simply involves a strong association between a measure and a reference indicator. It is not based on any implicit or explicit understanding of why the association exists. For example, if how quickly someone could snap their fingers 10 times was strongly correlated with their contracting HIV (which it isn't), then finger-snapping speed would be a criterion-valid indicator of HIV risk (again, which it isn't).
Construct validity involves assessing to what degree a measure
behavesin a fashion that would be expected for the construct it purports to represent. Unlike criterion validity, construct validity implies some understanding or at least hypothesis about why a new measure and another indicator should be correlated. Thus, it is theory-based. Often, this is assessed by laying out a series of predictions for how the construct in question would relate to other related phenomena, then observing the extent to which relationships between scores on the scale being evaluated and scores on relevant phenomena correspond to the predicted pattern. There are several subtypes of construct validity. Some of the most commonly discussed are:
- Convergent validity: the degree to which a measure correlates with another indicator when the variables they represent are known or believed to be related to one another. (e.g., we may expect scores on the scale measuring self-esteem to correlate with scores on the scale measuring confidence, because we understand those two variables to be related to one another.)
- Discriminant validity: the degree to which the construct of interest does not correlate with phenomena that it theoretically should not correlate with. (e.g., we expect low correlation between scores of a test of arithmetic skills and scores on social desirability.)
- Known groups validity: the degree to which a measure can differentiate between groups known to be similar or different with respect to a specific construct or classification. (e.g., we expect that patients with a diagnosis of anxiety disorder would score higher on an anxiety measure than people who have not been diagnosed with an anxiety disorder.)
As with reliability, it’s important to remember that high validity in one situation does not mean high validity in all situations. For example, an instrument that is excellent at measuring self-efficacy for HIV prevention in a population of adolescent girls in South Africa may be very bad at measuring self-efficacy for HIV prevention in adult male IV drug users in the urban United States. If this were the case, then the instrument would be valid in the first setting but not valid in the second. Looking at the literature and review process by which a scale was validated and carefully considering the similarities and differences between your population and the original study population will help you to select the most appropriate scale. It is also important to be sure that the variable as defined in any published validation studies is the variable in which you are interested.
- Obtain a copy of the instrument and source article.view
Obtain the instrument either on your own by following the links to the articles on the search results page or by emailing the CFAR Social and Behavioral Core at email@example.com to see if we have a copy of the instrument that we could share. Obtain the source article by following the links to the articles on the search results page or by entering the citation information into PubMed or Google Scholar.
- Compare the specific definition of what the scale captures to your definition of the variable in which you are interested. view
As an illustration, an instrument may be a valid measure of "anxiety" and you may have an interest in assessing "anxiety." However, if the measure is actually assessing public-speaking anxiety and the variable you wish to measure is anxiety about disclosing one's HIV status, then even if the measure proved valid in past research, it would not be a valid measure of the type of anxiety you wish to assess. Don't be misled by the names instruments are given. Look for evidence that they actually capture the variable that you wish to measure.
- Compare the populations the scale has been used in to yours.view
Another way to assess whether the instruments that you find in your search are right for your purposes is to see if they have been used with populations similar to the population you would like to study. To do this you can do your own search on how the measure has been used.
You may want to know whether a particular instrument has been used and tested with the population that you are interested in. To find the information, conduct a search using the name of the instrument AND the population of interest in Google Scholar or PubMed.
- Example: "HRQOL" AND "HIV positive" AND "women"
- Example: "Social Support Questionnaire" AND "HIV positive" AND "adolescents"
- Example: "Condom self-efficacy scale" AND "gay" AND "men"
If an instrument does not have a name, you may list the name of the author of the instrument AND a keyword indicating what the instrument intends to measure AND a population of interest.
- Example: "Hamilton" AND "Anxiety Rating Scale" AND "adolescents"
- Example: "Herek" AND "AIDS stigma" AND "HIV positive"
You may also find adaptations or modifications of the instruments (e.g., shorter or longer versions) with good reliability and established validity that you may want to use in your study.
Policy for Use of the Social and Behavioral Instruments (SABI) Database
This database and associated electronic resources are provided for researchers at University of North Carolina (UNC) Center for AIDS Research (CFAR), other research and educational institutions, and the general public for academic research and educational use. By using this website you agree to the following terms:
Information about SABI
SABI is a compilation of social and behavioral instruments intended to help researchers access and compare measures for use in HIV research. Content is collected from various journals under fair use and is subject to change as this database is being continually updated.
Utilization of the SABI database and associated resources is exclusively for educational or research purposes.
- Users must cite the original journal and author of all publications
- Users must reference SABI when reproducing or adapting database content
- All commercial use of resources found in SABI is prohibited
Other uses are permitted only with the express, written permission of the UNC CFAR Social and Behavioral Science Research Core.
UNC CFAR cannot and does not guarantee that the SABI database and associated resources are free of viruses, worms, infection, Trojan horses, time bombs, cancelbots or other computer programming routines intended to damage, detrimentally interfere with, surreptitiously intercept or expropriate any system, data or personal information, or that it is otherwise error-free.
You assume total responsibility and risk for accessing and using the SABI database through the internet.
You acknowledge and agree that any uploads or transmissions you make may be intercepted and used by an
unauthorized third party and that all of the risk associated therewith is solely yours. UNC CFAR makes
no express or implied warranties, representations or endorsements whatsoever (including, but not limited
to, warranties of title or non-infringement, or the implied warranties or merchantability or fitness for
a particular purpose) with regard to the SABI database and/or its contents. UNC CFAR shall not be liable
for any cost or damage arising either directly or indirectly from any use by you of the SABI database or
its contents. The work and all its contents are provided to you on an
as is and
as available basis.
You will hold harmless and indemnify UNC CFAR, its employees and agents, from any and all claims, suits, losses or damages that they may incur as a result of your misuse of the Work.
UNC CFAR reserves the right to revise the content of the SABI database or of these terms at any time.
For more information please email Catherine Grodensky, UNC CFAR Social and Behavioral Sciences Core Manager, at firstname.lastname@example.org
Register to create an account
Have you forgotten your password?