Professional Documents
Culture Documents
Diff Bet Exp & Non-Exp: randomization. – Cause & Effect. – manipulation of V. – assignment & distribution of V. –
control of V. – Use of treatment(manipulation). – distinction bet control & exp group.
TERMS
Dependent Variable (DV): criterion by which exp are judged. – V that is dependent on manipulation of IV. – is what is
being measured or observed. – the outcome that is expected to be caused by the IV.
Independent Variable (IV): V that is manipulated or altered independently of any other V. – hypothesized to be the
causal influence. – controlled to observe its effect on the DV. – should be clearly defined & operationalized. E.g., In a
study of the effect of diff study techniques on exam scores, the IV would be type of study used; condition 1: flash card
creation, condition 2: SQ3R, condition 3: Feynman techniques, condition 4: no technique (just reading). The DV
would be the exam score achieved by participants after using the diff techniques.
Experimental Group: group of participants that are exposed to the exp treatment e.g., the group that receives the 3
conditions.
Control Group: group of participants that are exposed to any treatment. – serve as reference point for comparison. –
helps account for any change that might naturally occur without the influence of IV. E.g., In same exp, this is the 4 th
group, where participants are not giving a specified technique just instruction to just read as per usual.
Randomization: assignment of subjects & treatments to groups based on chance in an unbiased way, ensuring that
each participant has an equal chance of being assigned to any group. – provided control by chance. – allows
assumption that groups are identical with respect to all variables except experimental treatment – eliminates bias –
eliminates the effect of the EV. E.g., In a large population of 1000 students, a researcher wants to select a sample of
100 students for a study. They use systematic sampling by first listing all students, then selecting every 10th student
from the list (10, 20, 30, and so on). This creates a systematic sample of 100 students.
Constant Error CE is error that occurs in the same experimental condition every time the basic experiment is
repeated that introduces an uncontrolled EV – a systematic bias. It causes measurement to consistently deviate in
the same direction. - Affects both precision and accuracy by introducing a consistent offset from the true value. It is
diff from random error which is an unpredictable variation in measurement in different directions without a specific
pattern. In a memory experiment, participants are asked to recall a list of words after a certain time interval. However,
the room temperature is consistently too cold. This cold environment affects participants' cognitive functioning,
making it harder for them to recall the words accurately. This systematic bias (constant error) introduced by the cold
temperature impacts the results of the memory experiment in the same way every time the experiment is conducted
under these conditions. E.g. In an experiment where the researcher is testing the effect of ambient lighting on
participants' mood, he measures the time participants spend solving a set of puzzles under different lighting
conditions. The constant error is introduced when the timer used consistently starts 5 seconds late for each trial
across all lighting conditions. Consequently, the recorded puzzle-solving times for all participants are consistently
longer than the actual time they took to complete the tasks, leading to a systematic distortion in the assessment of
the impact of ambient lighting on mood.
Sources of CE:
Experimenter bias: Effect on the subjects’ bhvr caused by an experimenter’s presence, actions, or comments. – is
the researcher’s influence on the outcome of the exp through his bhvr, preconceived notions, expectations or desires
about result. E.g., a ist testing a new anxiety reducing therapy strongly believes it is effective and may unknowingly
more encouragement, positive feedback or subtle cues or approval of patients receiving the treatment making the
making the subject feel more inclined to report improvement even if they didn’t feel it. In an experiment examining the
impact of praise on academic performance, the researcher, unintentionally influenced by personal beliefs,
consistently provides more encouraging feedback to participants in the experimental group. This experimenter bias
introduces a systematic distortion in assessing the true effect of praise on performance.
Guinea Pig Effect: Effect on experimental results caused by subjects changing normal bhvr or attitudes to cooperate
with experimenter. – occurs when subjects have knowledge on exp and change their response to the IV. E.g., ist
conducting study on social interaction in a controlled lab setting. Subjects might be more self-conscious because of
the knowledge that they are being observed leading to them changing their bhvr to fulfil researcher’s expectations. In
an anxiety study, participants who know they are being observed during stress-inducing tasks may exhibit heightened
anxiety levels due to the guinea pig effect, impacting the accuracy of the study's findings on natural anxiety
responses.
EXTRANEOUS VARIABLE: Variables other than the manipulated variables that affect the results of the experiment,
can invalidate the results. - can be natural characteristics of the participants, such as their intelligence level, gender,
or age, or features of the environment, such as lighting or noise. If left uncontrolled, extraneous variables can lead to
inaccurate conclusions about the relationship between the IV & DV. E.g., a researcher is conducting an experiment to
test the effect of caffeine on memory, the EV could be the amount of sleep the participants had the night before the
experiment. If the participants who had more sleep performed better on the memory test, it would be difficult to
determine whether the effect was due to caffeine or sleep. In a study on the impact of sleep on creativity, participants
solve puzzles after different sleep durations. Extraneous variable: breakfast. If the well-rested group coincidentally
enjoys a more substantial breakfast, their higher scores could be due to energy, not sleep itself.
Moderating Variable: A moderating variable is a factor that influences the strength or direction of the relationship between two other variables. In other
words, it affects the relationship between the independent variable and the dependent variable. A moderating variable often specifies conditions under
which a particular effect is stronger or weaker. For example, in a study examining the relationship between sleep quality and job performance, the
workload could be a moderating variable. The impact of sleep quality on job performance might be stronger for individuals with higher workloads
compared to those with lower workloads.
Mediating Variable: A mediating variable explains the process or mechanism through which an independent variable influences a dependent variable. It
helps in understanding the "why" or "how" of the observed relationship. In a hypothetical study on the relationship between exercise and improved
mood, self-esteem could be a mediating variable. The idea is that exercise might lead to improved self-esteem, and in turn, higher self-esteem might
contribute to an uplifted mood. The mediating variable provides insight into the underlying processes that connect the independent and dependent
variables.
Intervening Variable: The term "intervening variable" is less commonly used in contemporary research literature, and its meaning can sometimes
overlap with mediating variables. Generally, an intervening variable is a variable that comes between an independent variable and a dependent
variable in a causal chain. It's a variable that helps explain the relationship between the two. In practice, the distinction between mediating and
intervening variables may not always be clear-cut, and the terms are sometimes used interchangeably.
Controlling EV
Blinding: Technique used to control subjects’ knowledge of whether or not they have been given the experimental
treatment. Taste tests, placebos (chemically inert pills), etc. E.g., In a study testing a new pain reliever, some
participants receive the actual pain reliever while others receive a placebo. The participants don't know which
treatment they received, ensuring their expectations do not influence their pain reports.
Constancy of Conditions: Subjects in experimental & control groups are exposed to identical situations except for
differing conditions of the IV. E.g., In a memory experiment, all participants are tested in a quiet room with the same
level of lighting and temperature. This ensures that environmental conditions are consistent for both the experimental
and control groups, except for the variable being manipulated (e.g., different study techniques).
Order of Presentation If the experimental method requires that the same subjects be exposed to 2 or more
experimental treatments, error may occur due to order in which the treatments are presented. Method involves
changing the order in which diff lvls of IV are received by participants. – can only be done by counterbalancing.
Counterbalancing: ½ the subjects exposed to Treatment A first, then to Treatment B. Other ½ exposed to Treatment
B first, then to Treatment A. Eliminates the effects of order of presentation.
E.g., a study investigating the effects of two different learning techniques (Technique A and Technique B) on memory
retention. If the order of presentation is not controlled, it could lead to potential confounding effects.
Counterbalancing: Half of the participants are randomly assigned to Group 1. They first experience Technique A,
followed by Technique B. The other half are assigned to Group 2. They experience Technique B first, followed by
Technique A. E.g., Imagine an experiment testing the effect of music on memory. You want to compare classical
music to pop music, but you're worried that whichever music is heard first might have a stronger impression, skewing
the results. To counterbalance; 1. Randomly divide participants into two groups: Group A: Listen to classical music
first, then pop music. Group B: Listen to pop music first, then classical music. Within each group, ensure equal
numbers of participants experience each order: So, in Group A, half hear classical first and half hear pop first. The
same applies to Group B. Any potential bias from hearing one music first is equally distributed across both classical
and pop conditions.
VALIDITY
Internal: indicates whether the IV was the sole cause of the change in the DV. 3 conditions necessary for internal
validity: The cause and effect change together. The treatment occurs before changes in response variables. No
confounding or extraneous variable can explain the results of the study.
External: indicates the extent to which the results of the experiment are applicable to the real world i.e.,
generalizable.
Threats to Internal Validity (HCMTIDM)
History: Specific events in the external environment between the 1st & 2nd measurements that are beyond the
experimenter’s control. In a study on the effectiveness of a new teaching method, a major educational reform is
implemented halfway through the study, potentially impacting student performance. An experiment is conducted to
assess the impact of mindfulness meditation on stress levels among college students. However, during the course of
the experiment, the students have their final exams, which could independently affect their stress levels.
Cohort Effect: Change in the DV that occurs because members of one experimental group experienced different
historical situations than members of other experimental groups i.e., individ in a study have shared features due to
being in the group (cohort) making then differ from others. E.g., in a study on reading habits in diff age groups the
cohort effect is that older adults grew up in an era of physical books & teens with tech. This cohort specific diff in tech
& physical habits can influence study habits. E.g., Studying the cognitive abilities of individuals aged 60 and older to
assess the impact of technology use on memory and problem-solving skills. A cohort effect may occur due to the
participants belonging to different generations with distinct experiences related to technology. For example,
individuals who grew up with limited exposure to technology may exhibit different cognitive patterns compared to
those who have been immersed in digital technology throughout their lives.
Maturation: Effect on experimental results caused by experimental subjects maturing or changing over time. During
a daylong experiment, subjects may grow hungry, tired, or bored. In a child development study, researchers observe
improvements in children's problem-solving skills over time. However, this improvement may be due to natural
developmental processes rather than the specific intervention being studied.
Testing: The act of taking a pre-test affects participants' performance on subsequent tests. In before-and-after
studies, pretesting may sensitize subjects when taking a test for the 2nd time. May cause subjects to act differently
than they would have if no pretest measures were taken. E.g., In a memory study, participants may improve in the
post-test due to increased familiarity with the material, even if the intervention has no actual effect.
Instrumentation: Caused by a change in the wording of questions, in interviewers, or in other procedures used to
measure the dependent variable. E.g., In a study on anxiety levels, switching from self-reported questionnaires to
physiological measures of anxiety midway through the study may introduce inconsistencies. Or in an exp measuring
effect of a diet on weight loss. A study uses two different types of scales to measure weight loss among participants.
The difference in weight could be due to the difference in the scales rather than the effect of the diet plan
Selection: Sampling bias that results from differential selection of respondents for the comparison groups.
Systematic differences in characteristics between participants in different groups. E.g., In a study on the effectiveness
of a new educational program, the researcher inadvertently assigns more motivated students to the experimental
group. As a result, any observed improvements in academic performance might be due to the pre-existing motivation
levels of the participants rather than the program itself.
Mortality or Sample Attrition: Results from the withdrawal of some subjects from the experiment before it is
completed. Especially troublesome if some withdraw from one treatment group and not from the others (or at least at
different rates). E.g., In a long-term study on the effects of an exercise program on physical fitness, some participants
drop out of the study over time due to personal reasons. If those who drop out differ systematically from those who
remain in the study, it can introduce bias into the results.
Non-representative Sample: When the sample is not representative of the population of interest, the findings may
not apply to the broader population. E.g., In a study about political opinions, the sample consists of only registered
Democrats. The findings cannot be generalized to the entire population with various political affiliations.
Lack of Randomization in Sampling: Without random selection, the sample may not be representative, potentially
introducing bias. E.g., In a study on smartphone usage, participants are selected based on who volunteers. This non-
random approach may lead to an unrepresentative sample, as those who volunteer may have different smartphone
habits.
Using Non-probabilistic Sampling (e.g., Convenience, Snowball): Non-probabilistic sampling methods may
introduce bias, as they do not provide every member of the population an equal chance of being included. E.g.,
Conducting a survey on social media usage by posting a questionnaire on a social media platform. This method may
disproportionately include heavy social media users, potentially skewing the results and not representing the broader
population's bhvr.
OPERATIONAL DEFINITION: clearly defined set of procedures for measuring or manipulating the construct of
interest. It defines abstract concepts in concrete, measurable terms, allowing for consistent and objective
assessment. The key to an acceptable OD is that the procedure is specified precisely enough to allow replication by
others. Examples: quality of memory -- accuracy of recall in a certain task, depression -- Beck Depression Inventory
(survey) score, arousal -- galvanic skin response (conductivity of the surface of the skin).
HYPOTHESIS: a tentative statement to explain an observable occurrence. - A logically stated idea to show
relationship/compare variables (DV & IV). - Provides a clear & specific prediction or expectation. - Often indicates
expected direction of relationship (e.g., null or alternative). E.g., exposure to porn has no significant effect on sexual
bhvr of adolescents.
RESEARCH QUESTION: a broad open-ended inquiry about a topic that guides the research process. - Doesn’t
make specific prediction. - doesn’t show relationship bet variables. - More suitable for determining prevalence
(frequency of occurrence) or factors causing. E.g., what is the prevalence of aggressive sexual behaviour in teens. –
what are the factors responsible for masturbation in students.
RESEARCH DESIGNS
A strategic plan or procedure to be followed during a study to reach conclusions, with considerations given to the
participant selection & assignment to conditions, data collections & analysis. AKA research strategy & is a framework
used that includes methods & procedures to collect, analyse and interpret data.
Structure of research in
- Abstract
- Introduction: - Background – Scope – Statement of Problem – Operational Definition – Objective of Study
- Literature Review
- Method: Research Design (Exp D) – Population – Sample/ sampling technique – Instrument – Data collection – Data
analysis
- Result
- Discussion
- Conclusion & recommendation
- Reference
Cohort RD (QL or QT): design involves examining research subjects who have already been exposed to a research
topic. – common in medicin.e & is especially effective conducting ethical research on medical topics or risk factors. –
very flexible & applies to both secondary & primary data. – participants are selected based on the exposure status of
the individual. They are then followed over time to evaluate for the occurrence of the outcome of interest. Example:
Following a group of individuals born in the same year to study their career trajectories and achievements.
Historical RD (QL): design allows you to use past data to test your hypothesis. - research relies on historical data
like archives, maps, diaries, and logs. - useful for completing trend analysis or gathering context for a research
problem. - This method involves the systematic and objective location, evaluation, and synthesis of evidence to
establish facts and draw conclusions about past events. Historical research aims to show the importance of past
events in the present situation. Example: Analyzing primary documents and records to understand the political
climate during a specific historical event.
Survey RD (QL or QT ): Collects data from a sample of respondents using standardized questionnaires or
interviews. Allows for the collection of self-reported information types of surveys include:
• Interviews: Interviews are one popular type of survey. Interviews allow you to ask questions to a research
subject one-on-one, which can give you the opportunity to ask follow-up questions and gain additional insights.
• Online forms: You can also use online forms to conduct surveys. You can use many websites or software
programs to create intuitive online forms with a variety of question types, including short-answer and multiple-
choice.
• Focus groups: Focus groups are another key survey method. By using focus groups, you can facilitate
discussions with a group of research subjects to gain valuable research insights from your sample population.
• Questionnaires: Another type of survey is a questionnaire. In a questionnaire, you can simply list questions for
research subject to answer, making this an effective data collection method.
Systematic Review (QL): involves reviewing existing evidence and analyzing data from existing studies. – allows
use of previous research to come up with new conclusions. - as a review of the evidence on a formulated question
that uses systematic and explicit methods to identify, select and critically appraise relevant primary research, and to
extract and analyze data from the studies that are included in the review. E.g., : Conducting a comprehensive
analysis of published studies to assess the effectiveness of a specific medical treatment.
Meta-analysis RD (QT): uses a variety of populations from different existing studies i.e., method allows you to use
previous research to form new conclusions. A meta-analysis is a statistical analysis that combines the results of
multiple scientific studies. Meta-analysis is the use of statistical methods to summarize the results of
independent studies. Example: Combining data from multiple studies on a specific drug treatment to determine its
overall effectiveness.
MIXED-METHOD RD (BOTH QL & QT): combine multiple research methods to create the best path for a specific
research project. This type of research can include both qualitative and quantitative research methods. - wherein
researchers collect and analyze quantitative and qualitative data within a single study to answer their research
question. - help provide a more complete picture than a study that relies solely on quantitative or qualitative research.
. This is because it allows the researcher to gain a depth and breadth of understanding of a specific concept while
offsetting the weaknesses that are inherent when using either approach alone. E.g., Using surveys to collect
quantitative data on customer satisfaction and conducting interviews to gather qualitative insights on their
preferences.
Longitudinal RD (QL/T): design involves observing the same sample repeatedly over some time. This period might
be anywhere from a few weeks to several decades, depending on your particular research. - is mostly used in
medical research and other areas like psychology or sociology. – provides actionable insight. – often uses survey for
data collection that is either ql or qt. a survey creator does not interfere with survey participants. Instead, the survey
creator distributes questionnaires over time to observe changes in participants, behaviours, or attitudes. E.g., tracking
the physical and mental development of a group of children from infancy to adolescence.
EXPERIMENTAL DESIGNS
Designs where researcher manipulates one or more independent variables to observe their effect on a dependent
variable, while controlling for extraneous variables. They are used to establish cause-and-effect relationships
between variables.
Basic Keys
X = exposure of a group to an experimental treatment.
O = observation or measurement of the dependent variable. If multiple observations or measurements are taken,
subscripts indicate temporal order – i.e., O1, O2, etc.
R = individuals selected as subjects for the experiment are randomly assigned to the experimental groups.
Categories of Exp D
Pre-Experimental Quasi-Experimental True Experimental Designs
PRE-EXPERIMENTAL D
Don’t adequately control for problems associated with loss of internal or external validity i.e., they don’t control for EV,
don’t use random assignment, results can’t be generalised. – can’t be classified ad true experiment; this is due to the
lack of random assignment of subjects. – often used in exploratory research.
Merits: practical & feasible when true exp can’t be conducted. – provide preliminary insight & generate further
research hypothesis
Demerits: lacks rand ass. – lacks control over EV. – findings may be influenced by confounding. – can’t confidently
establish causality.
3 TYPES: One-Shot Design, One-Group Pretest-Post-test Design, Static Group Design after-only design with control
group
On e-Shot Design; after only design (no control)
A single group is exposed to an independent variable, and the effect on the dependent variable is measured once. - A
single measure is recorded from single group after the treatment is administered. - Study lacks any comparison or
control of extraneous influences - DIAG; X O1
Merits: Simplicity: The design is straightforward and easy to implement. Quick Results: Data collection and
analysis are relatively fast.
Demerits: Lack of Control Group: Without a control group, it's challenging to determine if changes observed are
truly due to the intervention or influenced by external factors. Internal Validity Concerns: The design is susceptible
to threats such as history, maturation, and testing effects. No Pre-Test: There is no baseline measurement before the
intervention, making it difficult to establish a cause-and-effect relationship. No rand ass
E.g., A researcher administers a new cognitive-behavioral therapy to a group of patients with depression and then
measures their mood using a self-report scale. E.g., Participants attend a stress reduction workshop, where they
learn relaxation techniques and stress management strategies. After the workshop, their stress levels are measured
using a standardized stress questionnaire.
One-Group Pretest- Post-test Design
A single group of subjects in the experimental group are measured before and after the treatment is administered. –
no control group for comparison. – no rand ass. - Offers comparison of the same individuals before and after the
treatment (e.g., training). - If time between 1st & 2nd measurements is extended, may suffer maturation. - Can also
suffer from history, mortality, and testing effects. Diag; O1 X O2
Merits: Within-Group Comparison: Allows for the examination of changes within the same group, providing a
baseline for individual progress. Control for Individual Differences: The design inherently controls for individual
differences, as each participant serves as their own control.
Demerits:
Lack of Control Group: Without a comparison group, it is challenging to attribute observed changes solely to the
intervention, as external factors may contribute.
History and Maturation Threats: Changes observed over time may be influenced by external events or participants
naturally maturing rather than the intervention itself.
No Random Assignment: The lack of a control group means there's no random assignment to treatment and control
conditions, limiting the ability to make causal inferences.
E.g., A researcher measures the self-esteem of a group of students before and after they participate in a mentoring
program. E.g., Exp; Assessing the Impact of a Cognitive Training Program on Memory Improvement, Pt; Participants
are assessed on their memory skills using a standardized memory test before engaging in any cognitive training. X;
Participants undergo a cognitive training program designed to enhance memory, involving various memory exercises
and strategies. Pst; After completing the training, participants are re-evaluated using the same memory test to
measure any changes in their memory skills.
Static Group Design; after-only design with control group [no pre-test]
Exp group is measured after being exposed to the experimental treatment. - Control group is measured without
having been exposed to the exp treatment. – no pre-measure is taken. – no rand ass. - Major weakness is lack of
assurance that the groups were equal on variables of interest prior to the treatment, this is caused by lack of rand
ass. Diag; Exp Group X O1
Control Group O2
Merits: Control Group Comparison: Enables researchers to compare the outcomes of the treatment group with a
control group, aiding in the assessment of the treatment's effectiveness. - It can be useful for testing the effects of
interventions that are difficult to manipulate or assign randomly.
Demerit: Major weakness is lack of assurance that the groups were equal on variables of interest prior to the
treatment, this is caused by lack of rand ass. - No Pretest Measures: Lacks pretest measurements for both groups,
making it challenging to account for baseline differences before the intervention.
E.g., researcher compares the academic performance of a group of students who received a mindfulness intervention
with another group of students who did not. E.g., Exp that investigates the effect of a stress management program on
employee well-being. The researcher selects two groups of employees from the same organization, one of which
receives the program and the other does not. The researcher then measures the well-being of both groups using a
questionnaire.
QUASI EXPERIMENTAL D
More realistic than true experiments because they capture complexity & variability of real world better than true with
its strict & artificial conditions. - Researchers lack full control over the scheduling of experimental treatments because
they study events or variations that are pre-existing or naturally occurring in groups. – No rand ass.
Merits; Useful when it is unethical to use rand ass & better to find people pre-exposed to treatment of interest. –
allows study in real world setting. – more practical & cost effective than true exp
Demerits; no rand ass. – less control over confounding. – limited internal validity.
TYPES; Time Series Design, Multiple Time Series Design: Same as Time Series but with control group
Time Series: Involves periodic measurements of the dependent variable over time. - After multiple measurements,
experimental treatment is administered naturally. After the treatment, periodic measurements are continued in order
to determine the treatment effect. is approach allows researchers to examine trends, patterns, and changes in
variables over time – no control group. – no rand ass. Diag; O1 O2 O3 O4 X O5 O6 O7 O8
Merits:
Rich Temporal Insight: Provides detailed information about changes and developments over time, allowing for a
nuanced understanding of the phenomenon.
Trend Identification: Enables the identification of long-term trends or recurring patterns that may be missed in cross-
sectional studies.
Forecasting Ability: Useful for making predictions or forecasting future trends based on historical data.
Demerits:
Data Intensity: Requires consistent and meticulous data collection over an extended period, which can be resource-
intensive.
External Factors: Changes in external variables, such as economic conditions or societal events, can complicate the
interpretation of trends.
Time Consuming: Longitudinal studies often require considerable time to yield meaningful insights.
E.g., researcher might collect data on the number of aggressive incidents in a psychiatric ward before and after the
introduction of a new treatment program. E.g., study the effect of a stress-reduction intervention on the daily mood of
college students. A researcher could measure the mood of a group of students every day for several weeks before
and after the intervention, using a standardized scale. The researcher could then use time series analysis to examine
how the intervention influenced the pattern of change in mood over time.
Statistical Design (Multiple Time Series); is considered when the experiment involves multiple levels of the factors
being considered. So, the periodic measurements in a specific time interval are taken from the event under
experimental conditions. - simultaneously analyzes multiple sets of time-ordered data. This design allows researchers
to explore relationships, patterns, and trends across different variables or groups over time. Multiple experiments are
conducted simultaneously to permit extraneous variables to be statistically controlled. This can be done by including
the extraneous variable as a covariate in the analysis, which helps to adjust the effect of the independent variable on
the dependent variable. – effects of the multiple IV is measured.
ADV; Can measure the effects of more than one independent variable. - Can statistically control specific extraneous
variables. - Economical designs can be formulated when each subject is measured more than once i.e it is
economical cause it does not require large sample size.
For example, suppose you are studying the effects of two different teaching methods on student performance. You
could conduct two experiments simultaneously, one for each teaching method. If there is an extraneous variable,
such as the time of day the classes are taught, that could affect student performance, you could include this variable
as a covariate in your analysis. This would allow you to adjust for the effects of this variable, reducing the likelihood
that it would confound the effects of the teaching methods on student performance.
Exp: impact of social media usage on mood flunctuation. participants: Two groups are recruited: a social media group
and a control group. Measurements: Mood is assessed daily for both groups over an extended period (e.g., 4 weeks),
using a self-reported measure like a rating scale. Intervention: The social media group maintains their usual social
media habits, while the control group significantly reduces their usage or abstains entirely.
Multiple Baseline and Intervention Phases: Both groups are observed for a baseline period before the intervention
(social media reduction) to establish individual patterns and control for potential seasonal or other cyclical effects.
Analysis: Statistical techniques like time series analysis are used to compare changes in mood over time between the
two groups. This allows researchers to: Identify if mood patterns differ between the groups before and after the
intervention. Investigate changes in mood variability within each group. Determine if the social media reduction led to
sustained or temporary changes in mood.
Factorial Design (mts); Used to examine the effects that the manipulation of at least 2 independent variables
(simultaneously at different levels) has upon the dependent variable. - The impact that each IV has on the DV is
referred to as the main effect ME. – DV may also be impacted by the interaction of the independent variables. This is
called the interaction effect IE.
Merits: Efficiency: Factorial designs allow researchers to study multiple factors and their interactions in a single
study, making efficient use of resources. - Generalizability: Findings from factorial designs can often be more
generalizable to real-world situations because they account for the complexity of multiple factors simultaneously.
Interaction Effects: Enables the examination of interaction effects, providing insights into how the combination of
factors influences the dependent variable.
Demerits:
Complexity: Factorial designs can become complex, especially as the number of factors and levels increases,
making them challenging to design and analyze. - Sample Size: Adequate sample sizes are required to detect
significant effects, and as the number of conditions increases, the needed sample size may become impractical.
Data Analysis Challenges: Analyzing factorial designs requires statistical techniques that can handle multiple
independent variables and interactions, which may be challenging for researchers without advanced statistical
training.
For example, a 2×2 factorial design would involve testing two factors, each with two levels. This would result in four
different conditions (2×2=4), representing all possible combinations of the levels of the 2 factors.
2. Conditions: Participants are randomly assigned to one of four conditions: Condition 1: Cognitive-Behavioral
Therapy, Short-Term, Condition 2: Cognitive-Behavioral Therapy, Long-Term, Condition 3: Psychodynamic
Therapy, Short-Term, Condition 4: Psychodynamic Therapy, Long-Term
3. Data Collection: Anxiety levels are measured before therapy, after therapy, and at follow-up sessions.
4. Data Analysis: Statistical analyses involve examining main effects of therapy type and duration as well as their
interaction effect on anxiety levels over time.
Grocery store chain wants to use 12 of its stores to examine whether sales would change at 3 different hours of
operation and 2 different types of sales promotions. Dependent variable is change in sales. Independent variables:
Store open 6 am to 6 pm. - Store open 6 am to midnight. - Store open 24 hours/day. Sales promotion: samples for a
free gift. Sales promotion: food samples. Called a 3 x 2 factorial design. Need 6 experimental groups (3 x 2 = 6)
TRUE EXPERIMENTAL D
A statistical approach to establish cause & effect. Characterized by presence of control & exp group & random
assignment.
Pretest-Posttest Control Group Designs; True exp design. - Experimental group tested before and after treatment
exposure. - control group tested at same two times without exposure to experimental treatment. - Includes random
assignment to groups. - Effect of all extraneous variables assumed to be the same on both groups, this is due to rand
ass which is said to neutralize the effects of EV making it same in both groups. Do run the risk of a testing effect.
Advantages: Allows for the assessment of causality by using a control group for comparison. - Controls for potential
pre-existing differences between groups through random assignment. - Provides a baseline measure of the
dependent variable before the intervention.
Limitations: May not be feasible or ethical in all situations, especially if a pretest could influence participants'
responses or if withholding treatment is ethically questionable. - Assumes that random assignment effectively
balances groups, which may not always be the case with small sample sizes.
Example: Imagine a study investigating the effectiveness of a new teaching method (independent variable) on
students' math scores (dependent variable).
Pretest: Both the experimental group and the control group take a math test before any intervention is introduced.
Random Assignment: Students are randomly assigned to either the experimental group (receiving the new teaching
method) or the control group (continuing with the current teaching method).
Introduction of Intervention: The experimental group receives the new teaching method, while the control group
continues with the current method.
Posttest: After a period of time, both groups take another math test to assess their performance.
Exp R O1 X O2
Control R O3 O4
Effect of the experimental treatment is (O2– O1) -- (O4 – O3)
Example
20% brand awareness among subjects before an advertising treatment
35% in experimental group & 22% in control group after the treatment
Treatment effect equals (0.35 – 0.20) – (0.22 – 0.20) = 13%
Most robust true experimental design. - Combines pretest-posttest with control group design and the posttest-only
with control group design. - Provides means for controlling the interactive testing effect and other sources of
extraneous variation. - Does include random assignment.
In this type of pretest posttest design, four groups are randomly assigned: two experimental groups E1/E2 and two
control groups C1/C2. Groups E1 and C1 complete a pre-test and all four groups complete a post-test. This better
controls for the interaction of pretesting and posttesting; in the “classic” design, participants may be unduly influenced
by the questions on the pretest.
Diagrammed as
Exp 1 R O1 X O2
Control 1 R O3 O4
Exp 2 R X O5
Control 2 R O6
Effect of independent variable (O2 – O4) & (O5 – O6) ▪ Effect of pretesting (O4 – O6) ▪
Effect of pretesting & measuring (O2 – O5) ▪ Effect of random assignment (O1 – O3)
E.g., A researcher wants to test the effect of a new online course on students’ academic performance. The researcher
randomly assigns 200 students to four groups: Group 1, Group 2, Group 3, and Group 4. Group 1 and Group 2 take a
pretest to measure their baseline knowledge of the course topic. Group 1 and Group 3 then enroll in the new online
course, while Group 2 and Group 4 do not. After the course is completed, all four groups take a post-test to measure
their learning outcomes. The researcher compares the posttest scores of the four groups to determine the effect of
the online course.
The Solomon Four-Group Design is considered robust for several reasons:
Controls Internal Validity Threats: This design can control threats to internal validity such as selection, history,
maturation, mortality, regression to the mean, and instrumentation.
Controls External Validity Threats: It can control threats to external validity like pretest sensitization, which a
standard two-group experimental design cannot control. Pretest sensitization occurs when the use of a pretest
increases or decreases the responsiveness of the participants to the study intervention.
Assesses Pretest Influence: The Solomon four-group design allows the researcher to check if the pretest
influenced the results. It also allows the researcher to test whether the pretest itself has an effect on the subjects.
Random Assignment of Participants: The assignment of participants is done at random, ensuring that each
participant has an equal chance of being placed in any group.
Simultaneous Measurement Across Groups: The outcome of interest is measured simultaneously across groups,
which helps maintain consistency and comparability of results.
Demerits
Resource Intensive: This design requires four research groups, which can be challenging in terms of resources and
time, especially in tightly funded research departments.
Disruption of Normal Practice: Most schools and organizations may not allow researchers to assign four groups
randomly as it can disrupt their normal practice.
Difficulty in Interpretation: Analyzing the data from four groups can be more complex than simpler designs, making
interpretation more challenging. There is a need for careful consideration and statistical expertise to draw accurate
conclusions.
Illustrate the S4-GD using A Hypothetical Exp
Topic: Effect of new teaching method on Math skills of SS1 students in Faith Academy
Hypothesis: H0- There is no significant effect of new teaching method on math skills of ss1 students of FA. H 1- There
is a significant effect of new teaching method on math skills of ss1 students of FA.
Draw S4-GD Diagram
Write out the formulas for testing diff effect
IV: New teaching method DV: Math skills
Sample: 20 SS1 FA students were randomly selected using systematic sampling. The class register was used as the
sample frame to get every 10th person from the class of 200 and assigned to 4 groups with 5 in each group. The 1 st 5
to exp 1, second 5 ctrl 1, third 5 to exp 2 & fourth 5 to ctrl 2.
Of the 4 groups 2 are exp & 2 are control.
Intervention: The treatment (new teaching method) would be administered to the 2 exp group while the control group
would receive standard teaching method.
Pretest: EXP1 & CTRL 1 receive a standard statistical quiz with a total score of 100
TREATMENT: EXP1&2
POSTTEST: EXP 1, CTRL1, EXP 2 &CTRL 2 receive another standard statistical quiz
Measurements
EXP1 (O1) EXP1 (O2) CTRL1 (O3) CTRL1 (O4) EXP2 (O5) CTRL2 (O6)
Pretest Posttest Pretest Posttest Posttest Posttest
48 69 59 58 73 47
55 70 42 54 69 54
45 66 57 45 75 44
59 66 42 50 72 50
41 69 42 48 77 40
Hypothesis
1. H0- There is no significant difference between the pre & posttest of experimental group 1 i.e., bet O1 & O2
H1- There is a significant difference.
2. H0- There is no significant difference between the pre & posttest of control group 1 i.e., bet O3 &O4
H1- There is a significant difference.
3. H0- There is no significant difference between the pretest of experimental and control group 1 i.e., bet O1 & O3
H1- There is a significant difference.
4. H0- There is no significant diff bet posttests of exp & ctrl group 1 i.e. bet O2 &O4. H 1- There is sig diff.
5. H0- There is no sig diff bet posttests of exp group 2 & control group 2 i.e., bet O5 & O6. H 1- There is sig diff.
6. H0- There is no sig diff bet posttest of eg1 & eg2 i.e., 02 & 05. H 1- There is sig diff.
7. H0- There is no sig diff bet posttest for cg1 & cg2 i.e., bet O4 & O6. H 1- There is sig diff
These hypotheses were tested by running these observations under SPSS t-test analysis, based on the decision rule
each hypothesis was either rejected or accepted. Decision rule: - If P ≤ 0.05 reject H 0, accept H1. – If P > 0.05 accept
H0, reject H1
Based on the analysis, the students in the experimental group who used the new teaching method showed a much
larger improvement in their math scores than the students in the control group who did not use the new teaching
method. This suggests that the new teaching method was effective in improving math skills. Therefore, I reject my
general null hypothesis and say that there is a significant effect of new teaching method on mathematical skills of
SS1 students of FA.
Non-Probabilistic Sampling: a method where researchers intentionally choose specific individuals or elements from
a population based on certain characteristics or criteria. It doesn’t rely on random selection & its goal is not to make
statistical inferences to the entire population but rather to gain insights into specific characteristics of groups of
interest.
- Convenience Sampling: adopted by researchers where they collect data from a conveniently available pool of
respondents. Ti is the easiest sampling method because participants are selected based on availability and
willingness to take part. Useful results can be obtained, but the results are prone to significant bias, because
those who volunteer to take part may be different from those who choose not to (volunteer bias), and the sample
may not be representative of other characteristics, such as age or sex. Note: volunteer bias is a risk of all non-
probability sampling methods. E.g., Surveying students in a class, interviewing shoppers at a supermarket
checkout.
Merits: Quick and inexpensive to implement. - Useful for pilot studies or initial data gathering. - Can provide
valuable insights for specific contexts.
Demerits: High susceptibility to bias, as the sample may not represent the larger population. - Limited
generalizability of findings. - Not suitable for quantitative research requiring statistical analysis.
- Quota Sampling: Researchers pre-determine quotas for different subgroups within the population and select
participants to fulfill those quotas. Mostly used by market researcher. E.g., Ensuring a sample with equal
proportions of male and female respondents, or different age groups. An interviewer might be told to go out and
select 20 adult men, 20 adult women, 10 teenage girls and 10 teenage boys so that they could interview them
about their television viewing. Ideally the quotas chosen would proportionally represent the characteristics of the
underlying population.
Merits: Increases the representativeness of the sample compared to convenience sampling. - Useful for studies
of diverse populations. - Can offer some degree of control over bias.
Demerits: - Requires prior knowledge about the population demographics. - Finding participants to fit specific
quotas can be challenging. - May still not eliminate bias, depending on how quotas are determined and filled.
- Judgement/Purposive sampling: aka selective or subjective sampling. this technique relies on the judgment of
the researcher when choosing who to ask to participate. Researchers may implicitly thus choose a
“representative” sample to suit their needs, or specifically approach individuals with certain characteristics. This
approach is often used by the media when canvassing the public for opinions and in qualitative research. E.g., a
researcher wants to study the buying patterns of high-end luxury car owners so he selects a sample of
individuals who they believe are most likely to purchase a luxury car.
Merits: Targeted approach ensures rich and in-depth data related to the research topic. - Useful for exploring
complex or sensitive topics. - Efficient for recruiting specific populations.
Demerits: Relies heavily on the researcher's judgment, introducing potential bias. - Limited generalizability
beyond the specific participant group. - May be time-consuming and resource-intensive to find suitable
participants.
- Snowball Sampling: commonly used in social sciences when investigating hard-to-reach groups. Existing
subjects are asked to nominate further subjects known to them, so the sample increases in size like a rolling
snowball. For example, when carrying out a survey of risk behaviours amongst intravenous drug users,
participants may be asked to nominate other users to be interviewed.
Merits: Reaches hidden or hard-to-reach populations effectively. - Creates a sense of trust and rapport with
participants. - Can uncover rich insider perspectives.
Demerits: Prone to bias, as the sample can become skewed based on the initial participants' connections. -
Difficulty in controlling for representativeness and generalizability. - Can be time-consuming and require
considerable effort to build trust and rapport.
Ethics in Exp
Ethics: the correct rules of conduct necessary when carrying out research. - a moral responsibility to protect
research participants from harm. ist have the duty to respect the rights & dignity of participants and must abide to
certain moral principles & rules of conduct.
Core Ethical Requirements
- Participation must be Voluntary: - all ethical research must be conducted with willing participants. – volunterrs
must not be coerced, threatened, or bribed into participation. This is especially important for researchers working
in uni & prison where stu & inmates are encouraged to participate in exp.
- Researcher must obtain Informed Consent (IC): IC is a procedure in which all study participants are told
about procedures and informed of any potential risks. Consent should be documented in written form. Informed
consent ensures that participants know enough about the experiment to make an informed decision about
whether or not they want to participate.
- Researcher must maintain participant confidentiality: Confidentiality is an essential part of any ethical
research. Participants need to be guaranteed that identifying information and individual responses will not be
shared with anyone who is not involved in the study.
- Discuss Intellectual property frankly: Intellectual property (IP) in research refers to the creations and outputs
that arise from research activities, it can be various things like discoveries, inventions, datasets & databases,
reports, etc. Owning and managing IP in research can be complex and involves multiple share holders like the
researcher, the uni or research institution, funding agency, etc. Academe's competitive "publish-or-perish"
mindset can be a recipe for trouble when it comes to who gets credit for authorship. The best way to avoid
disagreements about who should get credit and in what order is to talk about these issues at the beginning of a
working relationship, even though many people often feel uncomfortable about such topics.
- Be conscious of multiple roles: APA's Ethics Code says psychologists should avoid relationships that could
reasonably impair their professional performance or could exploit or harm others. But it also notes that many
kinds of multiple relationships aren't unethical--as long as they're not reasonably expected to have adverse
effects.
- Respect confidentiality and privacy: Upholding individuals' rights to confidentiality and privacy is a central
tenet of every psychologist's work. However, many privacy issues are idiosyncratic to the research population,
writes Susan Folkman, PhD, in "Ethics in Research with Human Participants" (APA, 2000). For instance,
researchers need to devise ways to ask whether participants are willing to talk about sensitive topics without
putting them in awkward situations, say experts. That could mean they provide a set of increasingly detailed
interview questions so that participants can stop if they feel uncomfortable.
Examples
1. Identification and Articulation of Research Problem:
Imagine the rising concern about information overload and its impact on student memory. We want to investigate whether spaced repetition learning, a
technique with spaced intervals between studying the same material, can improve long-term recall compared to traditional massed practice.
2. Title of the Experiment:
"The Effect of Spaced Repetition Learning on Long-Term Recall in Undergraduate Students"
3. Hypothesis:
Students who engage in spaced repetition learning will demonstrate significantly higher long-term recall of academic material compared to students
who use traditional massed practice.
4. Literature Review:
We analyze existing research on memory retention, exploring the theoretical benefits of spaced repetition and examining previous empirical studies
comparing spaced repetition to massed practice. This strengthens our research justification and informs our methodology.
5. Methodology: Research Design: We choose a pre-test/post-test control group design to assess the impact of spaced repetition. One group
(experimental) utilizes the spaced repetition technique, while another (control) utilizes traditional massed practice. Both groups complete pre-tests and
post-tests of academic material recall.
6. Sample and Sampling Technique:
We recruit 60 undergraduate students from various fields through random selection from a course enrollment list. This randomization reduces bias and
ensures generalizability of results.
7. Instrument and Test Validity:
We create a standardized test containing academic material relevant to the participants' studies. To ensure validity, we conduct a pilot test with a small
group and revise the test based on feedback.
8. Treatment:
Experimental Group: Participants receive access to a spaced repetition learning platform where they study the assigned material in intervals over
one week.
Control Group: Participants receive the same material and are instructed to study it massed (all at once) within one week.
9. Permission and Informed Consent:
We obtain approval from the university's ethics committee and ensure informed consent from all participants, explaining the experiment's purpose,
procedures, and risks.
10. Implementing Sampling and Randomization:
Using computer-generated random numbers, we assign participants to either the experimental or control group while ensuring equal group sizes and
balanced representation of various demographics.
11. Administering Treatment, Pretest, Intervention, and Posttest:
Pretest: Both groups complete the standardized test before any intervention.
Intervention: The experimental group uses the spaced repetition platform for one week, while the control group studies the material massed
within one week.
Posttest: Both groups complete the same standardized test one week after the intervention to assess long-term recall.
12. Measurements: We record individual scores on the pretest and posttest for each participant. These scores represent the dependent variable (long-
term recall). We also record demographic information as potential control variables.
13. Data Analysis:
Database: We store the data in a secure spreadsheet software.
Analysis: We conduct an independent samples t-test to compare the mean post-test scores of the experimental and control groups. This
determines whether the difference in long-term recall between the groups is statistically significant.
14. Results:
Let's assume the t-test analysis yields a statistically significant difference (p < 0.05) in mean post-test scores between the groups. The
experimental group, utilizing spaced repetition, demonstrates higher long-term recall compared to the control group employing massed practice.
15. Discussion, Conclusion, and Recommendation:
The results support our hypothesis, suggesting that spaced repetition learning is an effective strategy for improving long-term recall in undergraduate
students. We discuss potential limitations of the study and consider future research directions. We recommend incorporating spaced repetition
techniques into educational practices to enhance student learning and memory retention.
16. Reference:
Include a list of all research sources consulted throughout the study.