Professional Documents
Culture Documents
Finals Review
Session 1
Table of contents
• Factors that threaten internal validity: selection bias, history, maturation, testing,
instrumentation, regression to the mean, and experimental mortality.
o Inter-rater reliability: measures the consistency of results when different raters or observers rate the same behavior or event.
o Parallel forms reliability: measures the consistency of results when two versions of a test are administered to the same group
of participants.
o Internal consistency reliability: measures the consistency of results across items within a test.
Reliability
• Factors that threaten reliability: participant variability, measurement error, and experimenter bias.
• Enhancing reliability: can be achieved through strategies such as using standardized procedures,
training raters or observers, minimizing measurement error, and using multiple measures of the
same construct.
• Significance: Research reliability is essential for making comparisons between groups or across
time, as it ensures that any observed differences are not due to chance or measurement error. By
ensuring research reliability, we can have confidence in the validity of research findings, which is
crucial for making informed decisions and advancing knowledge in a field.
Validity vs. Reliability
Ensures that the conclusions drawn from Ensures that the results of a study are
a study are accurate and trustworthy. consistent and dependable.
→ The relationship between research validity and research reliability is complex. A study can be both
reliable and valid, or it can be reliable but not valid, or valid but not reliable. For example, a bathroom
scale that consistently measures weight with a high degree of precision is reliable but might not be valid
if it measures only the weight of one leg. On the other hand, a test that accurately measures a construct
but produces inconsistent results is valid but not reliable.
Validity vs. Reliability
Ensures that the conclusions drawn from Ensures that the results of a study are
a study are accurate and trustworthy. consistent and dependable.
• The relationship between research validity and research reliability is complex. A study can be both reliable
and valid, or it can be reliable but not valid, or valid but not reliable.
• In general, research validity is more important than research reliability. This is because a study that is not
valid is fundamentally flawed and cannot be trusted, while a study that is not reliable may still be useful.
Variable
• Definition: A variable can be defined as an attribute of a person or of
an object which “varies” from person to person or from object to object.
• Functions: Variables can be classified as dependent, independent, or
moderator variables. It is also possible to have intervening and control
variables as well
Dependent Variable
• Definition: The dependent variable is the variable which you observe and
measure to determine the effect of the independent variable.
Independent Variable
• Definition: The independent variable is the major variable which you hope to
investigate. It is the variable which is selected, manipulated, and measured by
the researcher.
Dependent vs. Independent Variable
Determine the dependent and independent variable in the following study:
Control Variable
• Definition: A control variable is a variable which is held constant in order to
neutralize the potential effect in might have on behavior.
Practice: Determine the variables
Determine the variables in the following study:
A study investigates the effects of explicit • Independent variable: explicit grammar instruction
grammar instruction on English language
learners' grammar skills. The study also
• Dependent variable: grammar skills
examines whether the effects of instruction are • Intervening variable: Students’ level of self-efficacy
mediated by students' working memory capacity.
• Control variable: Students’ prior English language
Additionally, the study holds constant students'
prior English language proficiency level. proficiency level
Practice: Determine the variables
Determine the variables in the following study:
Researchers are interested in whether the use of • Independent variable: use of peer feedback
peer feedback improves English language
• Dependent variable: English language learners'
learners' writing skills. The study also examines
writing skills
whether the effects of feedback are influenced
by students' cultural background. Additionally, • Moderator variable: Students’ cultural background
the study keeps students' prior English language • Control variable: Students’ prior English language
proficiency level at a constant level. proficiency level`
Quantitative Data Categories
1. Nominal – categories (e.g., Hindu, Catholic)
2. Ordinal – ranked categories (e.g., strongly agree, agree, undecided, disagree, etc.)
2. Ordinal – ranked categories (e.g., strongly agree, agree, undecided, disagree, etc.)
2. Ordinal – ranked categories (e.g., strongly agree, agree, undecided, disagree, etc.)
implies the ability to put data into rank order. Ordinal numbers convey more information,
in that their relative magnitude is meaningful.
2. Ordinal – ranked categories (e.g., strongly agree, agree, undecided, disagree, etc.)
2. Ordinal – ranked categories (e.g., strongly agree, agree, undecided, disagree, etc.)
Allows subtraction and addition but does not necessarily allow multiplication or division.
Length and weight are ratio scales.
Quantitative Data Categories
Triangulation
• It is good research practice to obtain data related to a given construct or
situation from a variety of sources.
• 4 types of triangulation
1. Variety of data source
2. Use of several evaluators
3. Use of multiple theories to interpret a set of data
4. Use of multiple method
Sampling
• Apparent conclusions of our research are
• less by the social reality under investigation
• more by the nature of the samples we use to collect data
→ Sampling is the basis of all research
Sampling
• Purpose: To use a relatively small number of cases to find out about a much larger number
• ‘Population’: group we wish to study
• ‘Sample’: group we actually involve in our research
• Representation:
• Sample drawn from a previously defined population
• sample being representative of this population
→ The researcher only sets out to generalise to the population from which the sample
was drawn
Sampling
1. Decide whether to use a sample, and why. 5. Choose a method of selecting population
4. Estimate the size of sample you need. compare with the ideal (or the population).
→ Snowball sampling
Non-probability Sampling
Determine which non-probability sampling method the following study used:
→ Quota sampling
Non-probability Sampling
Determine which non-probability sampling method the following study used:
→ Convenience sampling
Non-probability Sampling
Determine which non-probability sampling method the following study used:
A researcher wants to conduct a study on the experiences of English language teachers who use
online teaching platforms. She obtains a list of all English language teachers in the country and selects
20 teachers who have experience using online platforms and are willing to participate in the study.
What type of non-probability sampling method is being used in this study?
→ Purposive sampling
Questionnaires
• A questionnaire is a self-report instrument designed to gather ‘valid’ and ‘reliable’ information
o Not more than 30 minutes for adults (about 15 mins for mail questionnaire)
• Decide key demographic and classificatory variables [e.g., age, sex, education, occupation, residential
(or work) location, etc.]
• Wherever possible, avoid the use of negatively-worded questions/items; i.e., state all items in the POSITIVE
Principles of questionnaire design 5
• TRYOUT potential items with respondents drawn from the target
sample/population
Group Authors
• Spell out the full name of a group author in the reference list entry, followed by
a period: National Institute of Mental Health.
Thank you
&
Good luck