034 Chapter 3 Methodology 3.7 Limitations (a) Selection bias The aged care and community workers involved in this research volunteered to take part in the workshop. It is likely that at least some of these individuals had existing interest in the topic that influenced their decision to participate. Because individuals chose to participate and were not selected at random, the participants of this research are not representative of all aged care and community workers. This means the findings may not be generalised to apply to all workers involved in providing services to older adults in aged care and community settings. Further research is recommended to confirm these findings with larger, more representative samples. (b) Attrition bias The post-workshop survey, which was administered immediately following the workshop, had a very high response rate of 94.3%. The follow-up survey, which was conducted 2–3 months later, had a considerably lower response rate, with 57.3% of workshop attendees completing the questionnaire. Participant dropout is a common challenge in follow-up studies and raises the possibility of attrition bias, which occurs when there are systematic differences between participants who drop out of the study and those who continue.78 This may result in a biased sample and affect the generalisability of the findings. Several strategies were used to minimise attrition bias in this research, including: clearly outlining the requirements to participants from the outset; monitoring responses; sending follow-up and reminder emails; and providing incentives. (c) Social desirability bias Social desirability bias in surveys may occur when respondents are reluctant to admit to attitudes and behaviours that are considered socially unacceptable or undesirable.79 This may lead to respondents providing answers they believe are socially acceptable or desirable, rather than reporting their true beliefs or behaviours. To reduce the potential impact of social desirability bias, participants were reassured that their responses would remain anonymous and confidential. (d) Control group In research, a randomly assigned group of individuals who are not provided with the intervention under evaluation represents the counterfactual of what would have happened to participants without the intervention. In this research, the lack of a control group prevents us from drawing definite conclusions about the workshop as the cause of the significant differences observed between the preworkshop and post-workshop surveys. This limitation was partially addressed by having the same group of participants complete surveys before and after the workshop, thereby serving as their own control. As ageism is deeply ingrained and normalised, it is unlikely that substantial shifts in attitudes would occur spontaneously over a 2–3-month period. Additionally, participants’ own accounts of the changes that occurred because of the workshop offered rich and detailed insights and served to further support and validate the quantitative survey results.