Dated:1/12/2025
Time:2:00 pm to 3:41 pm
Research basics---class 1
Things discussed in class today:-
• Maintain a notepad to note down the things discussed during the class
• Be regular in attending
• Do what is expected of you. Go by what your guide says.
• Taking the student population is the most preferrable option but if you can
easily access and retrieve data from already available sample, then go for it
Arriving at a research question
1. Selecting a variable (could be something you have a query on or
curious about. [Link]).
2. After selecting a variable, find out how variable x affects variable y or
z . (to select other variables , you must understand the operational
definition of the variable you have chosen. Here in this
case…understand the operational definition of variable y).
3. Factors that affect the variable e.g. factors that affect grit .
Things to be noted: variables should be correlated, have an extensive
reading about the variable chosen, do enough literature reading.
4. Try to control and avoid confounding variables (external factor that
influences both independent and dependent variable ). [Link] we are
examining gender differences, we must ensure that the sample
includes an equal number of men and women.
5. A study should relate one variable to another and include
minimum three variables while planning your research study
SO after you have decided upon your variables(minimum 3) then
▪ Identify confounding variables
▪ Measure confounding variables to remove effects
6. Use a standardized questionnaire..(the scales must have reliability,
validity and norms or atleast make sure your scale is developed that is
..scale has reliability and validity)
7. Never bring variables that are highly correlated eg depression and
hopelessness. If two variables are too similar, they measure almost the
same thing — so studying both does not give any new or useful
information. Otherwise, it would distort the statistical measurement
because highly correlated variables overlap and make the analysis
unclear.
8. Variables must be selected carefully. Eg. Optimism and stress on
health. E.G. optimism has positive relationship on health, stress has
negative relationship on [Link] variables that relate to the
outcome and contribute useful information to the study
What should your scale measure?
A preview(maam explained the types of scale )
1. Continuous scale
Measures values that can take any number within a range. (e.g., height,
weight)
2. Discrete scale
Measures values in whole numbers only. (e.g., number of children)
3. Nominal scale
Only labels categories; no ranking or order.
4. Ordinal scale
Shows ranking or order
5. Interval scale
Has:
• Ranking
• Equal intervals between values
But no true zero( true zero means equal interval or represents complete
absence of what is measure)
Example: Celsius temperature
For meaningful statistical analysis, the scale should ideally be at the interval level
or higher, so you can measure how much each group differs from the average.
Summary so far:-
1. Chose your variables
2. Chose minimum 3 variables
3. Bring scales that measure each variables you
have chosen
4. Scales at score…chose a scale that gives
meaningful score
H.W.
Select variables(minimum 3)
Identify how your variables are related
Show studies done on these variables…..review of literature
Show scales along with their source, copyright (find out whether the scale is for
public domain
Or atleast authors contact details like email address)
While presenting all these to ma’am, make sure that you have everything opened ,
ready at hand,….
Framing research question
A research question should not be answered with just “yes” or “no”; it should
require explanation, analysis, and measurement..frame your research question in
such a way that there is NO yes/no answer to your question.
EXAMPLE- is there a relationship between stress, life satisfaction and sleep
quantity among university students?
(this question is too simple, answers only either yes or no)
But
“whether there is a relationship between stress, life satisfaction and sleep quantity
among university students ?”
(allows measurement and analysis..no clear cut answer)
Writing objectives based on your research
Main points maam explained….
---- When you frame objectives like:
• Relationship between Variable 1 and Variable 2
• Relationship between Variable 2 and Variable 3
• Relationship between Variable 1 and Variable 3
…you are repeatedly testing the same variables again and again in different
combinations.
This increases the chance of committing a Type I error because same variables are
being repeated.
Type 1 error is : rejecting null hypothesis when in reality null hypothesis is not
true
identify on whom your administering the scale onto
The objective must be framed accordingly
FIRST OBJECTIVE---- determining relationship between variables (x, y and z)
[testing multiple relationships between each variable,
{for ranking use spearman, for interval use pearson
correlation}]
SECOND OBJECTIVE---come up with the personal information of your
participants, that is the sociographic , socio-economic details like
name, age, gender followed by marital status family status,
income, number of family members, health etc ……(.details that
are needed to understand the study)…{also look into the order of
personal details to be given} basically factors that influence the
variables and formulate onjective accordingly …E.G. men and
women vary along perceived stress
Socio economic, socio-geographic details are taken to conclude—whether your
study is any different from previous study..eg;- previous studies show men has
much more perceived stress than women but your study concludes that women has
more perceived stress than men..
The second objective is to collect socio-demographic and socio-economic details
(age, gender, income, family status, etc.) to understand factors that influence the
variables and to compare whether our study differs from previous studies.
THIRD OBJECTIVE-- Find out how independent variable affect dependent
variables..that is whether variable x increase or decrease variable y or z
E.G Will high stress or low stress create difference on life satisfaction
FOURTH OBJECTIVE—IS regression---whether first 2 independent variables
affect the dependent variable
eg:-
To find out whether the two independent variables affect life satisfaction.
Hypothesis
Types of hypothesis..
1. null hypothesis: when research (existing) provides no relationship
---null always given or framed at population level to draw
inferences(inferential)
2. research hypothesis/ alternative hypothesis.. stated that there is a relationship
e.g. men and women vary in variables x,y and z
3. directional hypothesis- when one variable leads to another…when an
established theory states or predict how a variable would behave…
distributed on one side of tailed curve
4. non- directional :-states that there is a difference but doesn’t predict the
direction…….distributed on both sides of curve or 2 tailed test
here we will frame only one hypothesis , or incase of rising interventions ---
alternative hypothesis is framed accordingly
e.g. men and women do not differ in x,y,z (variables)
TO be CONSIDERED:-
---Formulate your topic uptill the hypothesis