Page 176 - Ana Mlekuž in Igor Ž. Žagar, ur. • Raziskovanje v vzgoji in izobraževanju: Izobraževanje učiteljic in učiteljev za raziskovalno učenje in poučevanje. Ljubljana: Pedagoški inštitut, 2024. Digitalna knjižnica, Dissertationes 48

raziskovanje v vzgoji in izobraževanju: izobraževanje učiteljic in učiteljev ...
ness and anxiety (e.g., Dambrun et al., 2019), we would assume practicing
mindfulness was also a protective factor during the period of high stress
(Zhu et al., 2021). For instance, mindfulness enhances acceptance and low-
ers intolerance of the uncertainty that was associated with higher levels
of generalised anxiety during the lockdown in the UK (Rettie & Daniels,
2020). The same study, however, established that coping strategies play an
important role. In future studies, it would therefore make sense to include
coping strategies to investigate their effect. The absence of significant ef-
fects may also be a specific characteristic of our sample. As our sample was
collected using a snowball method, allowing people (who were interested
in inner and body exercises) to apply for the study, it included a large pro-
portion of people involved in these kinds of activities, which is not the case
in the general population. The study should be replicated with larger and
more representative samples.

4.1 Limitations
The biggest limitation of our study was the small sample size, followed by
the additional attrition of participants throughout the assessments (how-
ever, this problem was addressed using a maximum likelihood algorithm
to assess the parameters of the LGC models), especially in the third as-
sessment, which may be a sign that many online studies were taking place
in Slovenia at that time, or of a loss of interest in the phenomenon. Snow-
ball sampling also does not ensure that the population sample is represent-
ative, and specific groups may be under-represented, such as in our case
men, and participants with lower levels of education. As the participants
responded at different times (e.g., participants were given access to the on-
line questionnaires on a weekly basis but they could respond to it later, and
there were time differences in T1 due to participants still deciding about
their participation in the study) the contextualisation of findings based on
week-to-week changes in restrictions is limited. One more limitation of
this study is not having pre-COVID-19 pandemic measures for compari-
son with the data. We must also take into consideration that the COVID-19
anxiety measure is new, due to the obvious novelty of the context, but has
shown adequate psychometric properties so far (Kozina et al., 2021). Al-
though the sample is very heterogeneous, our analyses did not include any
demographic variables as covariates, due to the biased sample, especially in
terms of gender and age. In future studies, we suggest expanding the size

176