STATISTICS FOR

EDUCATION B
Research designs, sampling techniques, and
data collection tools

Presented by: Sharon-Lee Engelbrecht

 LEARNING OUTCOME

Analyse relationships between variables to

identify trends and implications for classroom
practices.
 Introduction
Importance of research design and ensuring it shows valid and reliable results
Research design is a crucial component in conducting studies that yield valid and reliable results. It
serves as the blueprint for the entire research process, guiding how data is collected, analysed, and
interpreted.

Sampling techniques play a pivotal role in research design by determining how subjects or data points are
selected from the broader population. Choosing the right sampling method—whether random, stratified,
or convenience sampling—affects the generalizability and accuracy of the research findings.

Similarly, data collection tools are fundamental in shaping research outcomes. The selection of
appropriate tools—such as surveys, interviews, or observational checklists—ensures that the data
collected is relevant, precise, and comprehensive.
Research Design Overview
Definition: A research design serves as the enchanted framework for gathering and interpreting data.

Types of Research Designs:

Quantitative
(e.g., experimental, correlational, survey) – Gravetter et al. (2020), Frost (2019)
Quantitative designs focus on numerical data and statistical analysis, helping identify patterns and test hypotheses.
Experimental designs, for example, study cause-and-effect relationships by manipulating variables in controlled
settings.

Qualitative
(e.g., case study, ethnography, grounded theory) – Jhangiani et al. (2019)
Qualitative designs explore complex phenomena through detailed narratives, capturing participants' perspectives and
uncovering themes. Methods like ethnography provide cultural insights, while case studies analyze specific instances
deeply.

Mixed Methods
(blending quantitative and qualitative approaches) – Bless et al. (2006) Each research design has unique strengths suited
to specific questions.
Mixed methods combine quantitative and qualitative approaches, offering a comprehensive understanding of research
problems. This integration allows for nuanced insights that neither method could achieve alone.
Factors influencing design choice:
Research questions
Type of data needed
Ethical considerations

Example: How a classroom assessment study might require an

experimental/quasi-experimental design (Popham, 2017).

Next slide
 Using Experimental/Quasi-Experimental Design in Classroom

Case Study
A teacher wants to investigate whether formative assessments improve student performance in mathematics. Cause and Effect

Research Design Choice: Quasi-Experimental

Since students are already assigned to classrooms, random assignment is not possible, making a quasi-experimental design the most appropriate.

Two Groups
Experimental Group – Receives weekly formative assessments Data Collection
(quizzes, peer feedback, self-assessments). Pre-test at the start of the term to measure baseline knowledge.
Control Group – Follows the traditional teaching method without Post-test at the end of the term to measure improvements.
formative assessments. Student Feedback Surveys on assessment strategies.

Expected Outcome
If students in the experimental group show significant improvement, it suggests that formative assessments contribute to better learning outcomes
(Popham, 2017).

Why Quasi-Experimental?
Ethical & Practical Constraints - Schools can't randomly reassign students. It would be unethical to mix grades.
Real-World Application - Your findings can directly inform assessment strategies in education.
 Sampling Techniques
Definition: Sampling is the process of selecting participants from a population.
Which sampling method
ensures every member
Types of Sampling of a population has an
1. Probability Sampling (random, stratified, systematic) – Bless et al. (2006) equal chance of
selection?‌
Probability Sampling - involves methods where each member of the population has a
Choose participants based on
known, non-zero chance of being selected. This approach ensures that the sample is
$ 1,36 billion Total net assets
ease of access.‌

representative of the population, allowing for generalizations to be made about the Use random sampling.‌
entire group.
Select participants who are most
Random Sampling - This method gives every individual in the population an equal chance likely to give desired results.‌

350+ Projects
of being chosen. It is the most straightforward and fair approach, minimizing selection Submit‌
bias. ● Loading...‌

500+ Employee
Stratified Sampling - Here, the population is divided into subgroups or strata based on
shared characteristics. Random samples are then taken from each stratum, ensuring that

9.000 Clients
the sample reflects the diversity of the entire population.
Systematic Sampling - This technique involves selecting every nth individual from a list
of the population. The starting point is chosen randomly, but the process thereafter
follows a fixed, periodic interval.
 Definition: Sampling is the process of selecting participants from a population.

Types of Sampling

1. Non-Probability Sampling (convenience, purposive, snowball) – Jhangiani et al. (2019).

Non-Probability Sampling, on the other hand, does not provide every individual in the population with a
known chance of being selected. This method is often used when probability sampling is not feasible and
involves subjective judgment.
Convenience Sampling - Participants are selected based on their easy availability and proximity to the
researcher. While convenient, this method may not produce a representative sample, as it is prone to
bias.
Purposive Sampling - Also known as judgmental sampling, this method involves selecting individuals
who are deemed to have specific knowledge or characteristics relevant to the research purpose. It is
useful for targeted studies but limits the generalizability of results.
Snowball Sampling - This technique is often used for hard-to-reach or hidden populations. Initial
participants recruit others they know, creating a chain of referrals. While effective for accessing specific
groups, it can introduce bias due to reliance on social networks.
 Choosing the Right Sampling Method

Aligning sampling techniques with research goals.

Example: A school improvement study might use stratified random sampling (Bernhardt, 2018).

Next slide
 Case Study
A researcher wants to evaluate how different teaching strategies impact student performance across a school district.

Sampling Method: Stratified Random Sampling

Since schools have students from diverse backgrounds, the researcher divides the population into strata based on key characteristics:
Stratum 1: Grade levels (e.g., Grade 6, Grade 9, Grade 12)
Stratum 2: School type (e.g., public vs. private)
Stratum 3: Socioeconomic background (e.g., low-, middle-, high-income areas)

Study Process:
Each selected student’s performance is tracked before and after the introduction of a new teaching strategy.
Results from different strata are compared to determine which groups benefit the most

Why Stratified Random Sampling?

Ensures fair representation of all student groups. Data will represent real-time
Reduces bias by balancing socioeconomic and school-type differences. The researcher considers all factors.
Helps policymakers make data-driven improvements in education (Bernhardt, 2018).
 Definition: Instruments used to gather information.

Types:
Surveys & Questionnaires – Frost (2019)
Interviews (Structured & Unstructured) – Jhangiani et al. (2019)
Observations – Popham (2017)
Tests & Assessments – Gravetter et al. (2020)

Selecting the Right Data Collection Tool

What do you consider? Factors such as:
Nature of the research question
Participant accessibility
Validity and reliability

Example: In a behavioural science study, structured

questionnaires may be preferred (Gravetter et al., 2020).
Example - My Masters in psychology requires a structured
interview questionnaire
Ethical Considerations in Research

Informed consent, confidentiality, minimizing bias

Citing ethical principles from Bless et al. (2006), researchers are reminded of their responsibility to ensure that participants are
fully aware of the nature and purpose of the study before agreeing to take part. Informed consent is not merely a formality but a
fundamental ethical obligation that respects the autonomy and dignity of participants. This includes providing clear,
comprehensive information about the research, any potential risks, and the expected benefits, allowing participants to make an
informed decision about their involvement.

Confidentiality is another cornerstone of ethical research practice. Researchers must safeguard participants' privacy by ensuring
that their data is handled with the utmost care and discretion. This entails implementing robust data protection measures,
anonymizing data where possible, and ensuring that only authorized personnel have access to sensitive information.

Minimizing bias is crucial in maintaining the integrity of the research process. Researchers should strive to design studies that are
free from personal, cultural, or methodological biases, thereby ensuring that the findings are valid and reliable. This involves being
aware of potential sources of bias and taking proactive steps to mitigate them, such as using diverse participant samples and
employing objective data collection and analysis methods.

Citing the ethical principles outlined by Bless et al. (2006) serves as a valuable guide for researchers, emphasizing the importance
of conducting research with integrity, transparency, and respect for all individuals involved. By adhering to these principles,
researchers contribute to the advancement of knowledge in a manner that is socially responsible and ethically sound.
THANK YOU