org355: Student depression

1. What is the epidemiologic study design?

Asked how they felt over the previous two weeks, 37% showed “moderate to severe symptoms of depression” and 39% showed signs of “likely having some form of anxiety”.

2. Are these figure prevalence, prevalence odds, incidence risk, incidence rate?

The ONS cautioned that the statistics were “experimental”, based on a relatively small sample of about 2,000 first- and foundation-year students in English universities who were invited to take part via email.

3. What are the potential issues with sampling bias, and what are the potential implications of this bias on the study findings?

The survey also found that their satisfaction with life was significantly lower than that of the general adult population, at 6.6 out of 10 compared with 7.1, though similar to the general student population at 6.5.

4. How could the 'satisfaction' results be affected by confounding? What are some potential ways to address the issue?

5. What does "significantly lower" mean, statistically?

Re: Student depression

by | FATHIMA MINISHA - Friday, 29 October 2021, 6:55 AM

The hidden consequences of the pandemic that is slowly but surely emerging...:-( really scary...

1) The design seems like a cross sectional survey of first year students.

2) Considering that its a cross sectional survey, these are prevalence figures- not odds but prevalence (number with the condition-/total number surveyed)

3) So here they have a small sample of 2000 students who responded to the survey. In situations dealing with mental health issues, its more likely that ppl suffering from the mental health issue under study may per say not respond to the survey due to their condition. So, I am afraid these numbers could actually be an underrepresentation of the actual problem.

4) Well here they have compared college students with the general adult population- so age and stage of life are important confounders- older age group, who are working and earning more stable life, and of a different generation might actually have higher scores than college students... I think stratification is the way to go here you know- compare apples with apples rather than oranges.

5) Significantly lower here should be statistical difference- meaning the difference when compared using a hypothesis test gives a p value <0.05. Because clinically there does not seem to be a big difference between 6.6 and 7.1 out of 10...

What do you all think?

Fathima

Re: Student depression

by | JUDITH MARGARET BURCHARDT - Sunday, 31 October 2021, 5:56 PM

I agree with you on all points Fathima.

1. What is the epidemiologic study design?

Cross-sectional questionnaire study

2. Are these figure prevalence, prevalence odds, incidence risk, incidence rate?

Prevalence

The ONS cautioned that the statistics were “experimental”, based on a relatively small sample of about 2,000 first- and foundation-year students in English universities who were invited to take part via email.

3. What are the potential issues with sampling bias, and what are the potential implications of this bias on the study findings?

How were the students selected? Was there random sampling from a complete list of all first and foundation year students in England? If not, then were the students contacted representative of all first and foundation year students? If not then there is potentially selection bias in the study.

Did all students reply to the email? If not were those who replied different to those who did not, and in what way? Were students suffering with mental health problems more or less likely to reply than those who were not? Clearly if students with mental health problems were more likely to reply than those without then the study result will be biassed and it will appear that more students have mental health problems than is truly the case. The opposite is also true. This would be another example of selection bias.

There is also possibly information bias. A self report of symptoms of depression or anxiety was requested in the questionnaire. We do not know what these symptoms were, how severe they were or how common they are in the population. There was no validation of the answers given by individual participants and we do not know if the questionnaire was externally validated.

The survey also found that their satisfaction with life was significantly lower than that of the general adult population, at 6.6 out of 10 compared with 7.1, though similar to the general student population at 6.5.

4. How could the 'satisfaction' results be affected by confounding? What are some potential ways to address the issue?

Satisfaction with life might be confounded by age, socio-economic status, university, degree course and gender. It is possible to address this issue by stratifying responses by potential confounding variables to see what effect the confounders have on the exposure outcome relationship. If different strata of potential confounders give similar results then a summary estimate can be calculated. If different strata of potential confounders give very different results then these can be presented separately.

5. What does "significantly lower" mean, statistically?

It means that the 95% confidence intervals around 6.6, the estimate for student satisfaction levels, did not include 7.1, the population estimate.

Judith

Re: Student depression

by | FATHIMA MINISHA - Monday, 1 November 2021, 1:12 PM

Ah... I love that you prefer talking about CI and not p value.... I so want to get the p value out of my answers- hopefully i do better next time....;-)

Fathima

Re: Student depression

by | Hilja Eelu - Monday, 1 November 2021, 7:44 AM

1. What is the epidemiologic study design?

Cross sectional survey

Asked how they felt over the previous two weeks, 37% showed “moderate to severe symptoms of depression” and 39% showed signs of “likely having some form of anxiety”.

2. Are these figure prevalence, prevalence odds, incidence risk, incidence rate?

Incidence rate

3. What are the potential issues with sampling bias, and what are the potential implications of this bias on the study findings?

It could result in a differential misclassification bias (recall bias), in which we overestimate the relationship between students' mental health and being a new student during the COVID-19 pandemic.

4. How could the 'satisfaction' results be affected by confounding? What are some potential ways to address the issue?

Covid-19 could cause a decreased satisfaction of life, increasing the number of people with the outcome. It could be addressed by randomization or matching.

5. What does "significantly lower" mean, statistically?

If they ran a student's t test to test for differences in means, then the p value was likely smaller than 0.05.

Re: Student depression

by | MADHUTANDRA SARKAR - Sunday, 21 November 2021, 3:15 PM

1. This is a cross-sectional study.
2. These figures are prevalence figures.
3. Selection bias can occur because students who did not participate in the study may have different characteristics than those who do.
Information bias can also occur as some students suffering from the mental health issue may not respond.
Bias causes misleading interpretations of the study findings and false conclusions. Bias also limits the generalisability of the study findings.
4. Satisfaction results is affected by confounding by other factors which might affect the association between Covid pandemic and students’ mental health. The confounding factors might be age, gender, family status, etc.
This issue can be addressed by restriction. We can restrict our study to the students who are all in the same category of the potential confounder. We should include the students of same age or of same gender or of same family status, etc. only.
5. Statistically “significantly lower” means if we do a hypothesis test and calculate p-value, and p-value <0.05.