General principles of research methodology. Terms frequently used in this chapter. It is a course subject for fourth Pharm D in The Tamilnadu Dr.MGR. Medical University, Chennai.
2. •Research refers to a search for
knowledge, a scientific and systematic
search for pertinent information on a
specific topic.
• Research is an art of scientific
investigation or inquiry specially through
search for new facts in any branch of
knowledge.
By Gladys Stephen M.Pharm
3. OBJECTIVES OF RESEARCH
•To discover answers to questions through the
application of scientific procedures.
•1. To gain familiarity with a phenomenon or to achieve
new insights into it.
•2. To portray accurately the characteristics of a
particular individual, situation or a group
•3. To determine the frequency with which something
occurs or with which it is associated with something
else.
•4. To test a hypothesis of a causal relationship between
variables (such studies are known as hypothesis-testing
research studies).
By Gladys Stephen M.Pharm
4. • For doing research, statistics plays a major role. Many of the
terms that comprise statistical nomenclature that are familiar.
• Specifically, such terms are
• Variables and variations
• discrete and continuous variables,
• frequency distribution,
• population,
• sample,
• mean, median, standard deviation,
• variance, coefficient of variation (CV), range, accuracy,and
precision
By Gladys Stephen M.Pharm
5. VARIABLES
•Variables are the measurements, the values, which are
characteristic of the data collected in experiments.
•These are the data that will usually be displayed, analyzed,
and interpreted in a research report or publication. In statistical
terms, these observations are more correctly known as random
variables.
• Random variables take on values, or numbers, according to
some corresponding probability function.
•Duplicate determinations of serum concentration of a drug 1 hr
after an injection will not be identical no matter if the
duplicates come from (a) the same blood sample or (b) from
separate samples from two different persons or (c) from the
same person on two different occasions.
By Gladys Stephen M.Pharm
6. Continuous Variables
•A continuous variable is one that can take on any
value within some range or interval (i.e., within a
specified lower and upper limit).
•The limiting factor for the total number of possible
observations or results is the sensitivity of the measuring
instrument.
•When weighing tablets or making blood pressure
measurements, there are an infinite number of possible
values that can be observed if the measurement could
be made to an unlimited number of decimal places.
•Often, continuous variables cannot be easily measured
but can be ranked in order of magnitude.
By Gladys Stephen M.Pharm
7. Discrete Variables
•discrete variables can take on a countable number of values.
•These kinds of variables are commonly observed in biological
and pharmaceutical experiments and are exemplified by
measurements such as the number of anginal episodes in 1 week
or the number of side effects of different kinds after drug
treatment.
•Although not continuous, discrete data often have values
associated with them which can be numerically ordered
according to their magnitude. E.g rating scale.
•Discrete data that can be named (nominal), categorized into
two or more classes, and counted are called categorical
variables, or attributes; for example, the attributes may be
different side effects resulting from different drug treatments or
the presence or absence of a defect in a finished product.
By Gladys Stephen M.Pharm
8. Continuous Variables
•Continuous variables can always be classified into discrete
classes where the classes are ordered.
•For example, patients can be categorized as ‘‘underweight,’’
‘‘normal weight,’’ or ‘‘overweight’’ based on criteria such as
those listed in Metropolitan Life Insurance tables of ‘‘Desirable
Weights for Men and Women’’.
•Thus data can be classified as:
•1. Continuous (blood pressure, weight)
•2. Discrete (number of anginal episodes per week)
•3. Categorigical variable: categorical, ordered (degree of
overweight)
•4. Attributes: categorical, not ordered (male or female)
By Gladys Stephen M.Pharm
9. VARIATIONS
•Variation is an inherent characteristic of experimental
observations. To isolate and to identify particular causes
of variability requires special experimental designs and
analysis.
•Variation in observations is due to a number of causes.
For example, an assay will vary depending on:
•1. The instrument used for the analysis
•2. The analyst performing the assay
•3. The particular sample chosen
•4. Unidentified, uncontrollable background error,
commonly known as ‘‘noise’’
By Gladys Stephen M.Pharm
11. RATING SCALE
•In the assessment of pain in a clinical study of analgesics, a
patient can have a continuum of pain. To measure pain on a
continuous numerical scale would be difficult. On the other hand,
a patient may be able to differentiate slight pain from moderate
pain, moderate pain from severe pain, and so on.
•In analgesic studies, scores are commonly assigned to pain
severity, such as no pain 0, slight pain 1, moderate pain 2, and
severe pain 3.
•The scoring system above is a representation of a continuous
variable by discrete ‘‘scores’’ which can be rationally ordered or
ranked from low to high. This is commonly known as a rating scale,
and the ranked data are on an ordinal scale.
•The rating scale is an effort to quantify a continuous, but
subjective, variable.
By Gladys Stephen M.Pharm
12. Frequency Distributions
•The frequency distribution is an example of a data
summary, a table or categorization of the
frequency of occurrence of variables in various
class intervals.
•A frequency distribution of a set of data is simply
called a ‘‘distribution.’’
•For a sampling of continuous data, a frequency
distribution is constructed by classifying the
observations (variables) into a number of discrete
intervals.
By Gladys Stephen M.Pharm
13. •For categorical data, a frequency
distribution is simply a listing of the
number of observations in each class
or category, such as 20 males and
30 females entered in a clinical
study.
By Gladys Stephen M.Pharm
14. SAMPLE
•Samples are usually a relatively small number of observations
taken from a relatively large population or universe.
•The sample values are the observations, the data, obtained
from the population.
•The sample could consist of a selection of patients to
participate in a clinical study, or tablets chosen for a weight
determination.
•The sample is only part of the available data.
•Observations on a relatively small sample is done in order to
make inferences about the characteristics of the whole, the
population.
By Gladys Stephen M.Pharm
15. POPULATION
•The population consists of data with some clearly defined
characteristic(s). E.g. a population may consist of all patients
with a particular disease, or tablets from a production batch.
•The total available data is the population or universe. When
designing an experiment, the population should be clearly
defined so that samples chosen are representative of the
population.
•This is important in clinical trials, for example, where inferences
to the treatment of disease states are crucial.
•The exact nature or character of the population is rarely known,
and often impossible to ascertain, although we can make
assumptions about its properties.
By Gladys Stephen M.Pharm
16. Population Sample
Tablet batch Twenty tablets taken for content uniformity
Normal males between ages 18 and 65 years 24 subjects selected for a phase I clinical study
Sprague – Dawley weaning rats 100 rats selected to test possible toxic effects
of a NDC
Analysts working for company X Three analysts from a company to test a new
assay method
Persons with diastolic blood pressure
between 105 and 120 mmHg in the USA
120 patients with diastolic pressure between
105 and 120 mmHg to enter clinical study to
compare two antihypertensive agents
Serum cholesterol levels of one patient Blood samples drawn once a week for 3
months from a single patient
By Gladys Stephen M.Pharm
17. Population parameters
•Any measurable characteristic of the universe is called a parameter.
•E.g. the average weight of a batch of tablets or the average blood pressure
of hypertensive persons in the Tamilnadu are parameters of the respective
populations.
•Parameters are characteristic of the population, and are values that are
usually unknown to us.
By Gladys Stephen M.Pharm
18. Precision
•precision refers to the extent of variability of a group of
measurements observed under similar experimental
conditions.
•A precise set of measurements is compact
Accuracy
•Accuracy refers to the closeness of an individual observation
or mean to the true value.
•The ‘‘true’’ value is that result which would be observed in
the absence of error.
•E.g., the true mean tablet potency or the true drug content
of a preparation being assayed.
By Gladys Stephen M.Pharm
19. Bias
•Any deviation of results or inferences from the truth,
or processes leading to such deviation.
•Bias can result from several sources: one-sided or
systematic variations in measurement from the true
value (systematic error); flaws in study design;
deviation of inferences, interpretations, or analyses
based on flawed data or data collection; etc.
•There is no sense of prejudice or subjectivity implied
in the assessment of bias under these conditions.
By Gladys Stephen M.Pharm
20. • Accuracy can also be associated with the term
bias.
• The meaning of bias in statistics is similar to the
everyday definition in terms of ‘‘fairness.’’
• An accurate measurement, no matter what the
precision, can be thought of as unbiased, because
an accurate measurement is a ‘‘fair’’ estimate of
the true result.
• A biased estimate is systematically either higher or
lower than the true value.
By Gladys Stephen M.Pharm