2
Statistical Inference
Are there any circumstances where statistical inference would not be relevant in
assessing bio-statistical data?
Statistical inference is a procedure whereby inferences about a certain population
are made based on particular statistics computed from a sample of data taken from that
population. Statistical inference is utilized in biostatistics to assist in assessing the
magnitude and distribution of a certain phenomenon (
Casella & Berger, 2021)
. There are
circumstances whereby the statistical inference would be irrelevant in evaluating
biostatistical data. First, statistical inference may not be relevant in Exploratory Data
Analysis (EDA). The initial stages of data exploration emphasize data comprehension,
ascertaining patterns, and creating hypotheses. Therefore, statistical inferences will not
be the main goal at this phase. In this stage, visualization of data, descriptive statistics,
and other EDA approaches are utilized to understand more about the data without the
creation of formal inferences.
Besides, the statistical inference may not be relevant when the sample does not
represent the population. The purpose of various biostatistical studies is to conclude
about a particular population based on a smaller sample. Thus, whenever the sample does
not represent the population, generalizing of the results is compromised, making it
extremely hard to make accurate conclusions about the larger population. Also, samples
not representing the larger population are prone to sampling bias, whereby particular
sections are under or over-represented.
Lastly, statistical inferences may also not be relevant in case studies. In personal
case studies or research, the emphasis is on a comprehensive analysis of certain cases or
unique scenarios. Therefore, statistical inference may not apply since the objective is to
describe or examine personal cases comprehensively rather than generalize inferences
about the larger population.
How can you use statistical inference as part of the data analysis section of your
proposed bio-medical research project?
My proposed bio-medical research project is determining the prevention method
which is most effective against malaria transmission while recognizing that there are
certain parts of Ghana where those methods are inaccessible. I will use statistical
inference in the data analysis of my project in various ways. First, I will use statistical
inference to design an appropriate research and sampling strategy method. I will utilize
statistical approaches to examine the sample size essential to get enough statistical power
in order to identify significant differences between the different malaria prevention
methods. A cluster sampling method will be used to make sure there is a representative
from every region in Ghana.
Moreover, I will use statistical inferences to conduct regression analysis. In order
to control for possible confounding variables and ascertain the independent impact of
different prevention techniques, regression analysis will be utilized. For example, I will
use multiple regression to identify the relationship between different malaria prevention
techniques and the risk of malaria transmission while changing the appropriate covariates
like sex, age, gender, race, or tribe (Rossi, 2022). This will allow me to compute the
adjusted odds ratios (OR) and relative risks (RR), providing more sturdy and accurate
inferences.
