This paper presents the process of constructing a questionnaire to map out and assess the awareness and acceptance of diversity among staff working in healthcare institutions in Sweden. The intention of constructing the questionnaire was to illuminate the obstacles that may prevent a diverse healthcare institution from being sensitive to and congruent with diversity. The main strength of this study was to use various validity and reliability measures in constructing the questionnaire.
There is a shortage of standardized tools with high validity and reliability for assessing the level of awareness and acceptance of diversity in healthcare institutions
. One reason for this may be that the construction of reliable items and subscales to explore attitudes is more complex and difficult when it comes to sensitive topics such as diversity than is the case for constructing items about facts. Aspects such as ‘social desirability’ often become a problem
The first steps in the process of constructing a questionnaire is identification of items
[22–25], and the careful selection of them by conducting standardized tests, in regards to what the questionnaire is supposed to measure. One of the main concerns for assessing the individual’s perspective or attitude toward sensitive issues is that the individuals may be reluctant to express their “real” feelings. For minimizing this problem, the items needs to be selected carefully and formulated in the way the respondents are willing to answer them
[22, 25]. The selection of items as the most important part of the validity of the scale was the main focus of the first and second phases of this study. In this study, critical evaluation of content was made and the expert group was consulted in the process of item and sub-scale construction. As the questionnaire was revised, one of the most important concerns for the authors was to try to construct a questionnaire that was ‘user-friendly’; that is, easily understood by everyone in the target population. Furthermore, since the topics of the questionnaire were sensitive, involving moral issues, the risk of responding to the questions based on social desirability was taken into consideration. Most of the items and sub-scales refrained from investigating the behaviors or attitudes of the respondents themselves. Instead, the items posed questions that referred to the general situations and/or the respondent’s perceptions or experiences of the issues that she/he observed in her/his work.
Evidence based on response process generally comes from an analysis of individual responses and can contribute to questions about differences in meaning or interpretation of test scores among the relevant examinees. Written and spoken comments on the items in the first pilot study and the ‘think aloud study’ were analyzed and contributed to several changes in items and wording.
Several statistical analyses also were used to assess the reliability and validity of the instrument. These processes were multi-step endeavors and in each step specific analyses were performed with different sample settings to reduce the probability of bias. The results suggested that the questionnaire is valid and reliable to use in a Swedish context in the healthcare sector with dimensions regarding discrimination, interaction, stereotyping, and communication difficulties among healthcare staff and patients.
In relation to the issue of generalization of findings, the validity of the questionnaire and to what extent it can be used for utilization in different settings, it is important to consider to what degree evidence of validity based on test-criterion relations can be generalized to a new situation (for example, place, job, or education), without further study of validity in that new situation. The fact that the wording of the items is easy to understand and directed towards all categories of professions independent of their education level contributes to considering the questionnaire valid in many different settings.
It needs to be noted that a majority of the sample population in the final test were female, due to the nature of the sample population. In Sweden, almost 90% of nurses are female and the number of the nurses is higher than that of physicians
. We are aware that it can potentially increase the risk of bias and of validity of the questionnaire. We recommend additional testing of the questionnaire with equal numbers of men and women in order to avoid bias, if the questionnaire is to be used in other settings. We must also acknowledge that due to the lack of similar valid questionnaires/tools in Swedish, we could not examine the external validity of the questionnaire
[10, 27, 28].
Furthermore, since the topics of the questionnaire were sensitive and involved moral issues, the risk of social desirability was taken into consideration. The problem with sensitive questions and the reluctance of individuals to respond freely to items can be minimized by extensive and clear information dissemination to the informants about the purpose of the study, their rights to withdraw their participation, and the confidentiality of responses. Most of the items and sub-scales avoided investigating the behaviour or attitudes of the respondent herself/himself. Instead the items sought responses related to the general situations, i.e., discrimination and communications in the workplace and with patients and their significant others. Furthermore this study introduced a valid questionnaire for use in the healthcare sector concerning the awareness and acceptance of diversity, discrimination, interaction, and communication among staff and patients, and is not recommended for use in different study settings with different study perspectives. Also needing to be taken into consideration is the fact that the results from attitude measurement scales do not necessarily indicate that the respondents behave or act based on how they respond
[29, 30]. It needs to be noted that people act differently based on different circumstances, and several other social and environmental factors may shape individuals’ behaviors
. Thus we recommend studies in the future use this instrument with observations to validate to what extent people act based on how they respond. By doing this the researcher could get a clearer picture of the phenomena regarding the assessment of awareness and acceptance of diversity in healthcare sectors.