Quantitative Research Methods: Design & Sampling

• Lesson covers Research Methodology
• The lesson aims to enable students to:
  • Choose appropriate quantitative research design
  • Describe sampling procedure and sample
  • Construct an instrument and establish validity and reliability

Research Methodology Section Components:

• Research Design
• Respondents
• Instrumentation
• Procedure
• Data Analysis
• Potential Ethical Issues

Research Design

• In writing the research design section, one should supply the following:
  • Type of research approach
  • Type of research design
  • Description of research design (with citation)
  • Justification for the choice of research design
• Research approach refers to overall plan on how to study the chosen topic

Research Approach Types:

• Qualitative approach
• Quantitative approach
• Mixed-methods approach

Research Objectives and Study Nature in Different Approaches:

• Quantitative research aims for description, explanation, and prediction, studying behavior under artificial, controlled conditions
• Mixed-methods research uses multiple objectives, studying behavior in more than one context
• Qualitative research focuses on description, exploration, and discovery, studying behavior in its natural environment

Data Collection and Result Types

• Quantitative research collects numeric data via structured and validated instruments and aims for generalizable findings
• Mixed-methods research uses multiple forms of data collection and may provide corroborating findings
• Qualitative research gathers narrative data via semi- or unstructured interviews and aims for particularistic findings

Research Design Types

• Quantitative: Non-experimental (Descriptive/Survey, Correlational, Causal-Comparative), Experimental (Single-subject, True experimental, Quasi-experimental)
• Mixed-methods: Explanatory, Exploratory, Convergent (triangulation)
• Qualitative: Action, Case Study, Ethnographic, Grounded Theory, Historical, Narrative Inquiry, Phenomenological

Research Topics and Corresponding Designs

• Study on a COVID-stricken area in Taguig City - Case Study
• Attitudes of High School students on modular learning approach - Descriptive
• Impact of Google Classroom on teacher efficacy - Correlational
• Assessing student performance through Facebook rooms - Action
• Study on H1N1 flu in the U.S. in 2009 - Historical
• Effects of digital classrooms on student participation - Quasi-experimental
• Effect of new treatment procedure for covid patients - True experimental
• Lived experiences of a bereaved family due to COVID-19 - Phenomenological
• Healing practices of Bontoc tribal group against COVID-19 - Ethnographic
• Confidence Level of Selected Stakeholders in the opening of class - Convergent

Research Design Details

• A comprehensive short description/definition of the research design and justification of your choices should come from legitimate reference
• Example: A study at Universal College of Parañaque using a quantitative approach with a descriptive design to identify the use of smartphone as tool among Grade 12 students

Respondents

• In writing the respondents section, the following questions need answering
  • Who are they?
  • Where do they live? (locale)
  • What are the criteria for inclusion in the study?
  • Why them?
• Assignment of respondents to groups, and group assignments should be addressed in correlational or experimental study
• Sample size and its recruitment technique for probability/non-probability are necessary
  • Slovin's formula (quantitative)
  • Data saturation (qualitative)
  • (Links available for both)

Probability Sampling

• Probability sampling considers all elements and each element has an equal chance of being selected

Probability Sampling Types

• Simple Random
• Systematic
• Stratified
• Cluster

Non-Probability Sampling

• Non-probability sampling is based on ease of accessibility

Non-Probability Sampling Types:

• Convenience
• Judgmental sampling
• Quota
• Snowball

Data Saturation

• Data saturation refers to the point in the research when no new information is discovered in data analysis

Slovin's Formula

• n = N / (1 + Ne^2)
• n is the sample size
• N is the population size
• e is the margin of error
• 1 is a constant value

Cochran's Formula

• n。 = Z^2pq / e^2

Respondents Example

• A study will be conducted near Universal College of Parañaque located at Dr Arcadio Santos Avenue
• Grade 10 students will be the respondents because they will soon decide the strand to take in the senior high school
• With a margin of error at .05, a sample of 290 students will be obtained out of 1046 students following the Slovin's formula
• The sample will be obtained using simple random sampling where each student will be assigned of a number in a sampling frame and be selected using a random number generator

Instrumentation

• Instrumentation refers to the different ways and means of collecting data that will answer the research questions
• To construct an appropriate instrument, one must:
  • Read relevant literature
  • Decide the type of instrument;
  • Create instrument;
  • Seek expert advice;
  • Pre-test instrument (pilot-testing);
  • Revise instrument for final copy

Types of instrument:

• Adopted: from past researches used completely (no change)
• Adapted: from past researches used with changes that you make to fit your research objectives
• Researcher-developed: entirely constructed based on your research objectives and other theories/related concepts

Qualitative instruments

• Interview (face-to-face, phone, email, online)
• Observation (checklist, field notes, video/audio recording, photos & artifacts)
• Focused Group Discussion (FGD) (face-to-face, online)
• Document Analysis (official records: primary/secondary)

Quantitative instruments

• Scales (Likert, rating, semantic differential)
• Tests (IQ test, EQ test, Diagnostic test, Proficiency test, Achievement test, Aptitude test, Placement test, Psychological test)
• Survey (forms, questionnaires, checklists, inventories)

Instrument characteristics:

• It must contain the research title
• Must contain introductory statement, data-collection items, and a closing statement

Data collection items must be free from the following:

• Irrelevant, confusing, leading, double-barreled questions
• Questions using double-negatives

Expert Advice

• Constant consultation with research teacher and other related experts about your instrument will provide insights to develop and improve your instrument.

Pre-testing Instrument

• Pilot-testing is doing a practice data collection to a number of people; one can interview or ask them to evaluate your instrument in terms of:
  • Length of data collection
  • Clarity of directions
  • Ease of answering, and other problems/concerns
• Pilot-testing checks instrument validity correctness of data gathered
• Pilot-testing checks instrument reliability consistency of results based on combined answers

Revising Instrument for Final Copy

• Consider research teacher's suggestions and results of pilot-testing for editing and revising
• When drafting instruments section: be mindful of classification, the type, description, and the validity and reliability measures

Instruments may classify as:

• Researcher developed
• Structured questionnaire to be utilized in the study containing two parts:
  • Closed-ended
  • Multiple choice question asking for practical activities in EIM and a Likert scale question on the level of acceptability using PCBT in learning electrical wiring concepts
• The instrument will be face validated by the research adviser and content validated by a statistician
• The instrument will use parallel forms to determine its reliability

Validity:

• Content Validity: measures whether the instrument adequately measures the concept
• Face Validity: experts validate that the instrument measures what its name suggests it measures
• Criterion-related Validity: measures whether the instrument differentiates in a manner that helps to predict a criterion variable
• Concurrent validity: the instrument can predict a criterion variable in the present
• Predictive Validity: the instrument can predict a criterion variable in the future
• Construct Validity: the instrument taps the concept as theorized Convergent Validity: do two instruments measuring the concept correlate highly?
• Discriminant Validity: measure have a low correlation with a variable that is supposed to be unrelated to this variable

Types of reliability:

• Test-Retest
• Parallel Forms
• Inter-Rater
• Internal Consistency