Interpreting Statistics in Public Health Research Articles (STAT6000)

This report analyzes two research articles focused on public health statistics, fulfilling the requirements of a STAT6000 assignment. The first article, by Lam et al. (2014), examines alcohol and drug use at school leavers' celebrations, exploring hypothesis testing, sampling methods (convenience sam...Read More

Running head: PUBLIC HEALTH STATISTICS
Interpreting Statistics in Research Articles
Name of the Student
Name of the University
Author Note

PUBLIC HEALTH STATISTICS
Paper 1: Liam et al. (2014) - Alcohol and other drug use at school leavers' celebrations
1. The researchers present two aims for the study, which leads to development of two
separate hypothesis. The first hypothesis by the researchers states that the celebratory
events are associated with increased use of alcohol and other drugs and led to risky
behavioural engagements in the subject. The null would be that subjects are engaged
in higher alcohol and other drug use in celebratory events. The alternative hypothesis
would be that celebratory events has no statistical significance to show increase in the
use of alcohol and other drugs. The independent variable is celebratory event and the
dependent variable is alcohol and drug use (AOD).
The second hypothesis is concerned with the experiences of harm and the level of
alcohol and drug use. The null hypothesis states that higher the use of alcohol and
other drugs will lead to greater experiences of harm in celebratory events. The
alternative hypothesis would be that levels of alcohol and drug use has so statistical
association with the experiences of harm in celebratory events. The independent
variable is level of AOD use and the dependent variable is experiences of harm.
2. The researchers have used convenience or ‘opportunistic’ sampling method to engage
participants in the survey.
The primary advantages associated with this sampling method is the simplicity of
sampling which eases hypothesis generation and facilitates data collection in short
period (Jager, Putnick & Bornstein, 2017). Furthermore, it one of the most
inexpensive sampling method (Acharya et al., 2013).
The key disadvantages associated with this method of recruiting the participants is the
high vulnerability of selection bias and other influences, which may go beyond the
control of the researchers (Taherdoost, 2016). Furthermore, such sampling method is

PUBLIC HEALTH STATISTICS
associated with increase sampling error (Etikan, Musa & Alkassim, 2016). These
disadvantages reduces the credibility of convenience sampling method significantly.
3. According to the research study, majority of the respondents were 17 years (91%) and
the rest were 18 years old (9%). Over half of the population comprised of female
respondents (56%). The demographic characteristics of the sample population were
then analysed for descriptive such as hangover, vomiting, emotional outburst,
argument, injury and legal problems. The accommodation location, level of AOD use,
survey modality and gender were the key independent variables of which the study is
based upon.
4. The researchers assessed six factors using logistic regression analyses. The
researchers took in consideration six independent variables to measure the
experiences of harm and ability of safety strategies in attenuating those experiences.
The variables included mean quantity of alcohol consumed in an event’s single day,
the use of other drugs, safety strategies related to alcohol which were assessed using
PBSS (Protective Behavioural Strategies Survey), gender of the respondents, the
location of accommodation and the survey administration modality, whether it was
face-to-face or online. The researchers have utilised the Standard Drink (SD)
approach, which is a validated tool from the National Drug Strategy Household
survey. The drink specification, size of drink receptacle and the consumption number
of each type comprised the SDs, which the researchers used a basis of the study to
determine the alcohol consumption by the respondents.
5. The researchers initially stated that one of the negative consequences associated with
alcohol and other drug use is unprotected sex. According to the findings from the
research study, the prevalence of unprotected sex in the study population was 13.62
per cent and the odds ratio for significant unique association between unprotected sex

PUBLIC HEALTH STATISTICS
and the safest Protective Behavioural Strategies Survey used is 10.92 with 95%
confidence interval of [1.14, 104.23]. The factors controlled to compare with those
engaged in the highest frequency of safety strategies include the gender, other drug
use, accommodation location, survey modality and alcohol use. The results can be
interpreted to state that even with the greatest frequency of the safety strategies in
place, there is a 10.92 time more likelihood of report of unprotected sex. A unique
significant association with the safety strategy of greatest frequency and unprotected
sex is reported where P < 0.05. This showcases high statistical significance between
the two unique associations.
6. Reviewing the demographic characteristics of the study population, the sample can be
remarked as highly representatives of the national population of school students.
According to the sample description, it is stated that over 87% of the subjects from the
sample population were from independent school. Moreover, 91% of the subjects
were aged 17, which is the common age of the students during their final year of
schooling. The gender distribution of the population was also fairly more in number
of female students, which is compliant to the sex-ration of the particular age group in
Australia (Wasserstein & Lazar, 2016). However, geographical limitation can reduce
the generalisability of the research findings. The research study is limited to school
students who have attended celebrations, particularly school leaving celebration, on
the Rottnest Island, which is situated in the Western Australia, 20 km off the west
coast of Perth. Even though it is a popular location for events, the results may vary
with the variation in event celebration style in other parts of Australia.

PUBLIC HEALTH STATISTICS
Paper 2 – Wong et al. (2013) - The rising tide of diabetes mellitus in a Chinese population:
A population-based household survey on 121,895 persons.
1. The researcher's Wong et al. (2013) aimed at evaluating the prevalence of diabetes
that is self-reported with the help of household surveys along with an examination of factors
that are associated with diabetes independently. The study is focussed on exploring the
prevalence of self-reported diabetes, with no indication of what the result may look like.
Thus, reviewing the objective of the study, it is evident that no hypothetical statement can be
drawn from the primary objective of the study. However, few aspects that can be taken from
the original objective of the study to be rephrased into a hypothetical statement would be the
efficacy of household surveys and self-reporting in determining the prevalence of diabetes in
the Hong Kong region. The null hypothesis would be that self-reporting measures using
household surveys are adequate indicators of the prevalence of diabetes in a region. The
alternative hypothesis would be that self-reporting using household surveys does not help to
determine the prevalence of diabetes.
2. The researchers employed a stratified non-random sampling method to recruit
participants upon taking their information from the Census and Statistics department.
According to the descriptive statistics, the demographics characters of the study population
were the age, sex and the monthly household income. The first demographic character of age
was sub-divided into 8 age groups. The sex of the participants was either male or female, and
the monthly household income of the participants was further categorised into four bands of
income in Hong-Kong dollars.
3. The researchers conducted interviews with the study population in the year of 2001
(33,609 participants), 2002 (29,561 participants), 2005 (29,802 participants) and 2008 with
28,923 participants. Out of these total number of participants, 103,367 were aged above 15
years, and the average age of the sample population was 38.2 years. According to the

PUBLIC HEALTH STATISTICS
researchers, the survey conducted covered over 96% of the resident population from Hong
Kong and the response rate was 75% for the household surveys, consistent across all rounds
of the survey. The primary objective of the research study was to determine the prevalence of
self-reported diabetes using the method of household surveys. Reviewing the objective of the
study and the inferential statistics used in it, it can be stated that almost all the population of
Hong Kong was covered in the study, which increases the generalisability of the research
findings from this study substantially. These inferential statistics are accurate in ensuring that
the whole population of Hong Kong is considered and included in the study.
4. The researchers adjusted the prevalence of diabetes in the study population for age
and sex to determine comprehensive and reliable results. The age-adjusted prevalence of
diabetes in a sample study revealed sex-specific statistics for male and female population for
the four conducted studies. Age-adjusted prevalence of diabetes in male population for the
four studies were 2.80 % (2001), 2.87 % (2002), 3.32 % (2005) and 4.66 % for the fourth
survey in the year 2008. The age-adjusted prevalence of diabetes in female population was
3.25 % (2001), 3.37 % (2002), 3.77 % (2005) and 4.31 % for the fourth survey in the year
2008. The prevalence among male adults of all the diabetic population was 47.9% in 2008,
which is significantly lower than the same criteria for the female population at 69.3%.
Upon adjusting for sex to find the prevalence of diabetes according to the age-groups of the
population, it was found out that a drastic increase in the prevalence rate can be observed
with the increase in the age of the population. Progressive rising trend was observed between
the first conducted study and the study conducted in the year 2008.
5. The researchers analysed the independent association between the demographics of
the patients and the rates of diabetes with the help of multivariate regression analysis. The
adjusted odds ratio for older age groups was 32.2 [20.6-50.4 95% CI] for the age group of 40-
65 years with the significance of p < 0.001. The adjusted odds ratio for patients aged above

PUBLIC HEALTH STATISTICS
65 years was 120.1, with the significance of p < 0.001. Lower-income, that is, monthly
household income less than 9,999 Hong Kong dollar had adjusted odds ratio of 2.19. These
statistics indicate a progressive trend between the advancing age and prevalence of diabetes.
Thus, it can be remarked that with the increase in age, the prevalence of diabetes in a
population increases significantly. Moreover, the lower-income groups are indicated to be
highly vulnerable to diabetes and contribute majorly to the overall prevalence of diabetes in
the study population. On the other hand, the researchers have stated that there is no
significant association between the sex and prevalence of diabetes. The age and monthly
household incomes of the sample were the significant determiners to the overall prevalence
of diabetes.
6. The researchers have relied on self-reported data to conduct the study, which can be
regarded as one of the significant limitations of this study (Farshchi, 2017). In China, three
out of four diabetic patients are undiagnosed. Reviewing this fact, it can be indicated that
many diabetic patients who were unaware of their medical condition and may have presented
their actual health status when self-reporting, thus raising concerns on the credibility of the
data. Furthermore, several variables that can influence diabetic prevalence are unrecognised
by the researchers during their conduct of the research study. These variables include a
family history of diabetes, body mass index and lifestyle factors.
5.

PUBLIC HEALTH STATISTICS
References
Acharya, A. S., Prakash, A., Saxena, P., & Nigam, A. (2013). Sampling: Why and how of
it. Indian Journal of Medical Specialties, 4(2), 330-333.
Etikan, I., Musa, S. A., & Alkassim, R. S. (2016). Comparison of convenience sampling and
purposive sampling. American journal of theoretical and applied statistics, 5(1), 1-4.
Farshchi, H. R., Macdonald, I., Madjd, A., & Taylor, M. A. (2017). Benefits and limitations
of traditional self-report instruments. In Advances in the Assessment of Dietary
Intake. (pp. 1-17). CRC Press.
Lam, T., Liang, W., Chikritzhs, T., & Allsop, S. (2014). Alcohol and other drug use at school
leavers' celebrations. Journal of Public Health, 36(3), 408-416.
Jager, J., Putnick, D. L., & Bornstein, M. H. (2017). II. More than just convenient: The
scientific merits of homogeneous convenience samples. Monographs of the Society
for Research in Child Development, 82(2), 13-30.
Taherdoost, H. (2016). Sampling methods in research methodology; how to choose a
sampling technique for research. How to Choose a Sampling Technique for Research
(April 10, 2016).
Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: context, process,
and purpose.
Wong, M. C., Leung, M. C., Tsang, C. S., Lo, S. V., & Griffiths, S. M. (2013). The rising tide
of diabetes mellitus in a Chinese population: a population-based household survey on
121,895 persons. International journal of public health, 58(2), 269-276.