Research Methods Midterm Review Dr Dodge February 28
Research Methods: Midterm Review Dr. Dodge February 28, 2006
Introduction to Research
What is Scientific Research? u - “Scientific research is systematic, controlled, empirical, and critical investigation of natural phenomena guided by theory and hypotheses about the presumed relations among such phenomena. ” Kerlinger
The Scientific Method and the Empirical Approach The fundamental characteristic of the scientific method is empiricism (knowledge based on observations). u Most importantly, the scientific method embodies a number of rules for collecting, evaluating, and reporting data (observations put into measurable form). u
Goals of Scientific Research 1. 2. 3. To Describe Behavior To Predict Behavior To Determine the Causes of Behavior 4. To Understand or Explain Behavior - In order to accomplish these goals, scientific research can be conducted in ways that are both basic and applied
Basic vs. Applied Research Neither type of research is superior to the other u The progress of science depends on the synergy between the two u Both may cause the other to be modified u Both use the scientific method of research u
Evidence-Based Practice
A Hierarchy of Levels of Best Evidence Level 1 evidence: strong evidence from at least one systematic review of multiple well -designed randomized controlled trials. u Level 2 evidence: strong evidence from at least one properly designed randomized controlled trial of appropriate size. u Level 3 evidence: evidence from welldesigned trials without randomization, single group pre-post, cohort, time series, or matched case-controlled studies. u
A Hierarchy of Levels of Best Evidence u u u Level 4 evidence: evidence from welldesigned non-experimental studies from more than one center or research group. Level 5 evidence: opinions of respected authorities, based on clinical evidence, descriptive studies, or reports of expert committees Which level(s) should we strive for in our research?
Meta-Analysis Meta-analysis: method used to review research literature based on statistical integration and analysis of research findings u In meta-analysis, the DV is the effect size (i. e. , the outcomes or results of each study selected for review transformed into a common metric across studies) and the IV are study characteristics (i. e. , participants, interventions, and outcome measures). u
Meta-Analysis u Benefits of meta-analysis: • Synthesize the results from many studies succinctly and intuitively for nonscientific communities, • Illustrate the amount and relative impact of different programs on different criteria for policy decision-making purposes, and • Identify the most effective programs and highlight gaps or limitations in the literature to suggest directions for future research
Research Ethics
Areas of Focus in Research Ethics u. Harm u. Informed Consent u. Confidentiality u. Deception u. Reporting Results and Plagiarism
The Scientific Method
Science Testing ideas empirically according to a specific set of procedures that is open to public inspection u Based on objectively observed evidence u Without personal beliefs, perceptions, biases, attitudes or emotion u
The Scientific Method
The Scientific Method Step 1: Observation u Step 2: Form a hypothesis u Step 3: Use hypothesis to generate a testable prediction u Step 4: Design the study u Step 5: Conduct the study u Step 6: Perform hypothesis testing u
APA Style * * Refer to Dr. Pruett’s Power. Point!
Quantitative and Qualitative Approaches
Observing Behavior can be observed or measured. u Because behavior varies so much, scientists need to have numerous ways to measure and observe behavior in both experimental and non-experimental settings. u
Research Perspectives u Two major theoretical perspectives underlie much of social science research: • a. the positivist tradition, which “seeks facts or causes of social phenomena apart from the subjective states of individuals” • b. the phenomenological perspective, which “is committed to understanding social phenomena from the actor’s own perspective”
Research Perspectives Quantitative Research • generally stems from a positivist tradition u Qualitative Research • generally follows the phenomenological tradition u Both, in theory, are valid! u
Quantitative Research Quantitative research: Data are collected and observations are reported numerically. u Questionnaires, tests, and other measures to record frequency of behavior, occurrence of behavior and/or duration of behavior u After numerical data are collected, they are analyzed statistically. u
Qualitative Research: Data are collected from observations and interviews u Data often expressed in non-numerical terms using language and images – analyzed systematically u Summaries of discussions, interviews, and video- or audio-taping of behaviors u
Quantitative vs. Qualitative Design Quantitative Research Designs: Experimental Quasi-Experimental Ex post facto Correlational Descriptive Time Series/Single Subject Qualitative Research Designs: Ethnographic Case study Phenomenological Historical Philosophical Grounded Theoretical
Quantitative vs. Qualitative Summary Neither is superior to the other u Choosing one or the other is based on what is needed in your study u High quality research projects may incorporate aspects of both approaches. u
Introduction to Qualitative Research * * Refer to Dr. Simpson’s Power. Point!
Introduction to Statistics
Statistics and Research Design u u u Statistics: Theory and method of analyzing quantitative data from samples of observations in order to help make decisions about hypothesized relations. Statistics are merely tools used in research design! Research Design: Plan and structure of the investigation to answer the research questions (or hypotheses)
Statistics u There are two types of statistics • Descriptive Statistics: involves tabulating, depicting, and describing data • Inferential Statistics: predicts or estimates characteristics of a population from a knowledge of the characteristics of only a sample of the population
Statistics & Parameters • Parameter: a value, usually unknown (estimated), used to represent a certain population characteristic • Statistic: a quantity that is calculated from a sample of data. Used to give information about unknown values in the corresponding population.
Descriptive Statistics • Nominal scales u No numerical or quantitative properties. A way to classify groups or categories. u Gender: Male and Female • Ordinal scales u Used to rank and order the levels of the variable being studied. No particular value is placed between the numbers in the rating scale. u Restaurant Ratings: 4 Stars, 3 Stars, 2 Stars, and 1 Star
Descriptive Statistics • Interval scales u Difference between the numbers on the scale is meaningful and intervals are equal in size. NO absolute zero. Allows for comparisons u Temperatures on a thermometer • Ratio scales u Scales that do have an absolute zero point than indicated the absence of the variable being studied. Can form ratios. u Weight: 250 pounds is ½ of 500 pounds
Descriptive Statistics u Frequency Distributions • Constructed by summarizing data in tables according to the number or frequency of observations in each category, score, or score interval • Data can be concisely summarized in bar graphs, histograms, or frequency polygons
Descriptive Statistics u Measures of Central Tendency • Mode u Most frequently occurring score • Median u Score that divides a group of scores in half. • Mean u Preferred whenever possible and is the only measure of central tendency that is used in advanced statistical calculations u An average of all scores. Add up scores and divide by total number of scores.
Descriptive Statistics u Measures of Variability • Range u Calculated by subtracting the lowest score from the highest score. u Used only for Ordinal, Interval, and Ratio scales • Variance u The extent to which individual scores in a distribution of scores differ from one another • Standard Deviation u The square root of the variance u Most widely used measure to describe the variation of a set of observations in a distribution.
Descriptive Statistics u u Correlation • Statistical summary of the degree or magnitude and direction of the relationship or association between two variables • Correlations can be positive or negative Linear Regression • Make predictions on a new sample of observations from the findings on a previous sample
Inferential Statistics: Sampling u u u Degree to which those surveyed are representative of a specific population Sampling frame: set of people who have the chance to respond to the survey External validity: degree to which the sample frame corresponds to the population to which the researcher wants to apply the results
Inferential Statistics: Sampling Two basic types: probability and non -probability u Probability sampling: including random sampling u Non-probability sampling: including convenience sampling u
Response Rates u u Most government-sponsored surveys require response rates of 75% A response rate of 70% is very good, 60% is good, and 50% is adequate Post-cards, follow-up letters, and telephone calls are used to increase the response rates Results of non-response bias can be examined by comparing those who respond early with those who respond after follow up
Questions?
- Slides: 41