VARIABLES AND DATA

BIOSTATISTICS AND EPIDEMIOLOGY LECTURE

CLASSIFICATION OF DATA
Z
1. Independent
- X, Exposure, Predictor, Explanatory
- Factors that would elicit an effect on your dependent Y X
variable.
b. MEDIATING VARIABLE
- Independent —> Dependent
- Independent eliciting effect on the Z variable
- Will elicit the outcome of the Y
2. Dependent
- Indirect
- Y, outcome, response
Z

3. Covariates
- Confounding,
Y X
- Can affect the X and Y variable separately
- False associations
Z c. MODERATING VARIABLES

Y X Z

Y X
TYPES OF COVARIATES
a. CONFOUNDING VARIABLES CLASSIFYING VARIABLES
- Indirect
1. Quantitative variable
- Confounding affecting the independent variable thus - derived from actual measurements of an object,
affecting the dependent variable
respondent or an event
- Confounding affects the dependent - Quantitative data
Ex.
X -Grandmother’s smoking during pregnancy 2. QUALITATIVE VARIABLE
Y -Grandchild’s Asthma
- facilitate the characterization of an object, respondent
Z -Maternal Age, Parity, Education, Yearly Salary, Pre- pregnant or an event
BMI, Maternal asthma, Child’s Birth weight.
- Qualitative data

1|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
SCALES OF MEASUREMENT

INTERVAL
a) MEASUREMENTS
CONTINUOUS - Assignment of numbers to objects or events according
RATIO to set of rules
QUANTITATIVE

b) INTERVAL
DISCRETE ORDINAL
- Quantitative scale wherein the difference between
DATA two variables has meaning
DISCRETE ORDINAL - known distance between 2 measurements
- has no true zero value
QUALITATIVE - does not reflect the absence of a variable
NOMINAL
- ex. Time, Temperature
CATEGORICAL
-
ORDINAL c) RATIO
- has a true zero value
- indicates the absence of a variable
TYPES OF DATA - Consistent interval relationship
- Unit or interval refers to a constant amount of the
1. CONTINUOUS variable throughout the measurement range
- Data is measures and has a value within a range - Consistent ratio relationship
- ex. Rainfall in a year, tumor size - The proportions of the magnitudes of variables are
accurately reflected by the proportions of numerical
2. DISCRETE scores.
- Numerical data that has a finite no. Of possible values.
- Whole numbers d) NOMINAL
- Ex. No. Of employees in the office - classifying two different categories
- Categories must encompass all individuals
3. CATEGORICAL - Categories must be mutually exclusive
- Data can be sorted according to group or category - Frequencies of occurrence of the classes
- ex. Types of products sold - *Dichotomous variables- 2 groups

2|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
e) ORDINAL
- ranking of the data
- consists of rank ordering TYPES OF VARIABLES
- represents rank CATEGORICAL Frequency Table
(nominal/ordinal) Cross Table
- does not reflect magnitude
Bar chart
- Numerical data with statement Pie chart
Mode
f) DISCRETE ORDINAL QUANTITATIVE CONTINUOUS (ratio/interval) Histogram
- SEVERITY or magnitude Scatterplot
- ex. Pain from 0-9 Mean (normal distribution)
Median (skewed distribution)
Min
g) ORDINAL QUALITATIVE
Max
- no magnitude and limited choices Range
Standard deviation

NOMINAL ORDINAL INTERVAL RATIO

“order” of Yes Yes Yes CLASSES OF STUDY DESIGNS
values
“counts” or Yes yes Yes Yes STUDY DESIGNS
“frequency of - METHODOLOGY used to collect information to address a
distribution”
MODE research question
Yes Yes Yes Yes
MEDIAN yes Yes Yes
MEAN Yes Yes Main goal:
Can quantify Yes Yes - attain the highest degree if objectivity
the difference - prevents biases and opinions and values that affects our
between each results
value
Can add or Yes Yes
subtract value Concern:
Can multiply Yes - Oversimplification
or divide
value
Operational definition:
Has true value Yes

3|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
- Specification of the manner of measurements of the
variable and how variables would be employed in the
study

LEVELS OF ABSTRACTION STUDY DESIGNS

I. THEORY
- ideas for abstract statements
- ex. Poverty leads to poor health

CASE SERIES
II. CONCEPTS
- poverty and poor health
CASE CONTROL
III. INDICATORS
- phenomena or factors that would lead to the existence
of a concept OBSERVATIONAL COHORT
- ex. Low income, way of living, diseases
CROSS
IV. VARIABLES SECTIONAL
- Components of indicators that can be measured
- Ex. Prevalence incidence, income, no of people in a HISTORICAL
COHORT
household

V. VALUES STUDY DESIGNS RANDOMIZED

- The actual MEASUREMENTS of the variables CONTROLLED
TRIALS
- EX. 15 people NON
RANDOMIZED

SELF-
EXPERIMENTAL
CONTROLLED
SEQUENTIAL

META- ANALYSIS CROSS OVER

NO CONTROLS

4|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE

EXPERIMENTAL
- involve and intervention and interest lies in the effect the
intervention has on study subjects

SAMPLING TECHNIQUES

SAMPLING
- measuring a small portion of something and then
making a general statement about the whole thing

5|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
- same ratio

SAMPLE SIZE WHY DO WE NEED SAMPLING?

- Number that is adequate enough to answer research √ SAMPLING makes possible the study of a large,
questions but not so much to involve too many subjects heterogeneous population
- statistical, clinical, and practical considerations √ Sampling is for economy
√ Sampling is for speed
SAMPLE PROCEDURES √ for accuracy
- dependent on the characteristics of the population from √ saves the source of data from being all consumed
which the sample is to be drawn from
- major error in public health studies
*DEFECTIVE SAMPLING
SAMPLING TERMS AND CONCEPTS - If sampling is biased, not representative or too small, the
1. STUDY POPULATION conclusion may not be valid and reliable.
- Organisms, objects or events of a specified type; where - If characteristics of the sample are not common, then
the sample is actually selected. the conclusions might be erroneous
- If the population is very large and there are many
2. SAMPLE sections and subsections, then sampling procedure
- people who are actually selected to participate or to becomes very complicated
be the subject of the study - If the researcher does not possess the necessary
technical skills in sampling procedure, then sampling
3. SAMPLING UNIT becomes biased and unrepresentative
- element or set of elements considered for selection

4. SAMPLING FRAME
- actual list of sampling units from which the sample, or
some stage of the sample is selected

5. REPRESENTATIVE SAMPLING
- if aggregate characteristics of the sample closely
approximate the aggregate characteristics of the
population

6|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
- selection based on situation

QUOTA A. Accidental
- samples are selected when the researcher meets them
NON- by chance.
ACCIDENTAL
PROBABILITY - disadvantage: does not represent the entire population
well. Lack of representativeness
CONVENIENCE - employed when all sampling techniques are no longer
available
PURE
SAMPLING
RANDOM B. Quota
- specified number of persons of certain types are
SYSTEMATIC included in the sample
- there is representation but it is not proportional
STRATIFIED
PROBABILITY C. Convenience
RANDOM
- process of picking out people in the most convenient
PURPOSIVE and fastest way

2. PROBABILITY
CLUSTER - Proportional to the pop.
- System of. Selection
- equal chance of participation

A. Pure Random
TYPES SAMPLING - simplest form
- everyone in the population has an equal
1. Non Probability
- sample generated is not proportional to the population B. SYSTEMATIC
- no system of selection

7|Page
CUTE
VARIABLES AND DATA
BIOSTATISTICS AND EPIDEMIOLOGY LECTURE
- Sample members from a larger population are selected
according to a random starting point but with a fixed,
periodic interval. This interval, called the sampling
interval, is calculated by dividing the population size by
the desired sample size. As long as the list does not
contain any hidden order, this sampling method is as
good as the random sampling method.
- Often used instead of random sampling,
advantageous than random sampling technique
because of its simplicity

C. STRATIFIED RANDOM
- Involves the division of a population into smaller sub-
groups known as strata. In stratified random sampling or
stratification, the strata are formed based on members'
shared attributes or characteristics such as income or
educational attainment. Random samples are then
selected from each stratum.

D. PURPOSIVE
- Non-probability sample that is selected based on
characteristics of a population and the objective of the
study. Purposive sampling is different from
convenience sampling and is also known as
judgmental, selective, or subjective sampling.

E. CLUSTER
- With cluster sampling, the researcher divides the
population into separate groups, called clusters. Then,
a simple random sample of clusters is selected from the
population either by simple random or systematic
random sampling. . The researcher conducts his
analysis on data from the sampled clusters.

8|Page
CUTE