You are interested to explore if arsenic exposure at baseline of the study is associated with Attention Deficit
Hyperactivity Disorder (ADHD) in children after 5 years of the baseline. State the null and alternative
hypothesis for examining this relationship.
Identify three categorical and three continuous variables in the SPSS dataset and indicate whether they are the
indicators of environmental exposure or biomarkers (biological/physiological indicators) or sociodemographic
characteristics. Use a table to present your answers.
This study has initially followed a cross-sectional design and later it became a prospective cohort study after
collecting data on intellectual disability and ADHD outcomes among the participants. State the advantages that
a prospective cohort study would offer as opposed to a cross-sectional study. Review your textbook to find the
answers.
We would like to estimate if there is an association between arsenic exposure in childhood and postadolescence ADHD among the study participants. Draw the 2X2 table that will allow you to calculate the OR
and RR by hand (just show the table; calculation is not necessary). Use the following two dichotomous
(categorical) variables – i) “ArsenicExposureCategory” as the exposure or predictor variable; and ii) “ADHD” as
the health outcome variable.
Calculate OR and RR with p-values and 95% Confidence Intervals for showing the association between
arsenic exposure in childhood and post-adolescence ADHD among the study participants. Use SPSS to
compute OR and RR.
For question 5, do you find p-values and 95% CI for both OR and RR statistically significant? Why or why not?
Briefly explain your findings.
Do you see any change in OR when you take the two confounding variables maternal education and sex (boys
vs girls) into account at the same time while computing OR? How much is the change? Show the adjusted OR,
corresponding p-value, and 95% CI. What is the objective of calculating adjusted OR for the relationship
between arsenic exposure in childhood and post-adolescence ADHD?
Using variables that are listed in the SPSS datafile as examples, state three potential sources of the following
random errors (give just one example for each of the following random errors):
Poor precision
Sampling error
Variability in measurement
Do boys and girls differ significantly for the following health outcomes? Conduct independent sample t-tests
and state your findings (you can copy the results from SPSS).
Body Mass Index (BMI) – show unit in the results
TPO Antibody – show unit in the results
Neurobehavioral test TAP preferred
What external validity issue do you see in the Bangladesh Adolescent study (hints: internal or external validity.
Sample Solution