Joni Lakin David Shannon Margaret Ross Abbot Packard Auburn University Auburn University Auburn University University of West Georgia
Figure: http://www.auburn.edu/~jml0035/eera_chart.pdf Presentation slides: http://www.auburn.edu/~jml0035/eera_pres.pdf
What is the question you re trying to answer? What are the variables of interest implied by the question? How will you measure the variables of interest?
What kind of question are you asking? Descriptive Relational Comparison Do you have 2 or more than 2 variables? Do you have 1 or more than 1 independent variables? 1 >1 Are the groups related or independent? Are the groups related or independent? 2 >2 Independent Related Independent Related Nominal What scale are the variables of interest? Frequency Counts; Percentages; Mode What scale are the variables of interest? What scale are the variables of interest? Contingency Multiple logistic coefficient (Phi) regression What is the scale of the dependent variable? Chi square What is the scale of the dependent variable? McNemar test (2 groups)or Cochran Q test (>2 groups) What is the scale of the dependent variable? Chi square What is the scale of the dependent variable? Cochran Q test Ordinal Interval/ Ratio Mode; Median; Range Mode; Median; Mean; Range; SD; quartiles Rank order coefficient Productmoment correlation coefficient Multiple regression Multiple regression Mann Whitney U test Independent samples t test or One Way ANOVA (between Wilcoxon signed ranks test or Friedman ANOVA Paired samples t test or One way ANOVA (within subjects) Kruskal Wallis ANOVA Factorial ANOVA (between subjects) Friedman Two Way ANOVA Factorial ANOVA (within subjects)
Descriptive Research Purpose Determines and describes the way things are Compares how sub groups view issues and topics Importance Frequently used in research studies Used to influence opinion Useful for investigating a variety of educational problems Obj. 1.1, 1.2, 1.3
Differences from Qualitative Research Descriptive Research Uses observations and interviews Studies the topic from the researcher s perspective Qualitative Research Uses observations and interviews Studies the topic from the participant s perspective Obj. 1.4
Classifications of Descriptive Research Classified by how data are collected Self report Individuals respond to statements or questions about themselves Observation Data is collected by the researcher watching participants Obj. 3.1 & 3.2
Survey Data Collection Methods Surveys represent the most common type of self report measures Questionnaires Interviews Advantages of conducting surveys Less time is required Less expenses are incurred Larger samples can be used Obj. 4.1
Conducting Survey Research State the problem The topic must be of sufficient significance to motivate a respondent and justify the research effort Select the participants Must be able to provide the desired information Must be willing to participate Obj. 4.2
Conducting Survey Research Constructing the questionnaire Appearance Visually attractive use of white space Brief Carefully planned content and item formats Clearly identify sub topics Use structured items if possible Use responses that can be checked or circled rather than written Obj. 4.5
Conducting Survey Research Methods to collect data Mailed surveys Advantages efficient, inexpensive, easily standardized, easy to score, anonymous or confidential Disadvantages low response rate, inability to probe E mailed surveys Advantages quick, efficient, inexpensive, easily standardized, easy to score Disadvantages can only assure confidentiality, respondents might not have access to e mail Obj. 4.4
Conducting Survey Research Telephone surveys Advantages high response rates, efficient Disadvantages requires lists of telephone numbers, requires training Personal administration Advantages efficient if respondents are in close proximity to the administrator, allows for probes Disadvantages time consuming, expensive, requires training Personal interview Advantages rich complete responses Disadvantages little standardization, takes time to administer, expensive Obj. 4.4
Descriptive Statistics Purpose to describe or summarize data in a parsimonious manner Basic types of descriptive statistics Central tendency Variability Objective 2.4
Central Tendency Purpose to represent the typical score attained by subjects Three common measures Mode Median Mean Objective 4.1
Central Tendency Mode The most frequently occurring score Appropriate for nominal data Median The score above and below which 50% of all scores lie (i.e., the mid point) Characteristics Appropriate for ordinal scales Doesn t take into account the value of each and every score in the data Objectives 4.2, 4.3, & 4.4
Central Tendency Mean The arithmetic average of all scores Characteristics Advantageous statistical properties Affected by outlying scores Most frequently used measure of central tendency Objectives 4.2, 4.3, & 4.4
Variability Purpose to measure the extent to which scores are spread apart Four measures Range Quartile deviation Variance Standard deviation Objective 5.1
Variability Range The difference between the highest and lowest score in a data set Characteristics Unstable measure of variability Rough, quick estimate Objectives 5.2 & 5.3
Variability Variance The average squared deviation of all scores around the mean Characteristics Many important statistical properties Difficult to interpret due to squared metric Objectives 5.2 & 5.3
Variability Standard deviation The square root of the variance Characteristics Many important statistical properties Relationship to properties of the normal curve Easily interpreted Objectives 5.2 & 5.3
Relationships Among Variables Correlation and Regression
Pearson r Ratio scale (continuous) Spearman rho Kendall s Rank Order Ordinal data Point Biserial Continuous scale dichotomous Phi Coefficient Dichotomous variables Level of Measurement
Bivariate Correlation What is the relationship between level of constructivism in teaching and student motivation? What is the relationship between belly dancing skill and years of musical training? (Tabachnick & Fidell) Examples
Just because something correlates does NOT necessarily mean one variable causes another. Thumb size and reading ability in elementary level students Predict relationships based on theory Causality
Theory Higher levels of constructivist teaching (predictor or independent variable) leads to or predicts higher levels of student motivation (dependent variable). Questionable (in my opinion) More years of musical training leads to higher levels of belly dancing skill. Simple Regression: One Variable Predicts Another
Theory Do higher levels of constructivist teaching and higher levels of student motivation predict higher levels of student achievement? Belly Dancing Skill Years of musical training Body flexibility Multiple Regression: Two or More Variables Predict the Dependent Variable
Path Analysis training Age flexibility Belly Dancing Skill Adding Complexity
1. Measuring the Variables constructivism Motivation achievement 2. Questions 1 What is the relationship between level of constructivism in teaching and student motivation? 3. Question 2 Do higher levels of constructivist teaching and higher levels of student motivation predict higher levels of student achievement? Flow Chart
Tabachnick, B. G., and Fidell, L. S. (2007). Using Multivariate Statistics, 5th ed. Boston: Allyn and Bacon. Pedhazur, E. J. (1997) Multiple regression in behavioral research, 3 rd ed. Florence, KY: Cengage Learning. Pedhazur, E. J. & Schmelkin, L. (1991) Measurement, Design, and Analysis: An Integrated Approach. Hillside, NJ ; Lawrence Erlbaum Associates. Bruce Thompson s Home Page: http://people.cehd.tamu.edu/~bthompson/ Regression Resources
* Two Independent Groups
* *Dependent *Growth mathematics EOG test *Independent *Males Single-sex 5 th grade math *Females Single-sex 5 th grade math *Analysis *Chi-Square
* *Dependent *Decoding multi-syllable words *Independent *Instruction in backwardchaining method *Analysis *Mann-Whitney U Test
* *Dependent *5 point Likert scale Survey *Independent *General education teachers *Special education teachers *Analysis *Independent t-test
* *Dependent *Scholarship of Engagement scores *Independent *Lectures *Instructors *Assistant professors *Associate professors *Professors *Analysis *One-Way ANOVA (between subjects)
* Two Related Groups
* *Dependent *Recidivism *Independent *Levels of change in scores in self-efficacy Analysis McNemar test
* *Dependent *Difference between during and after internship of stress levels *Independent *General education student teacher *Special education student teacher Analysis Wilcoxon signed ranks test
* *Dependent *Pre-&Posttest Stress scores *Independent *General education student teacher *Special education student teacher *Analysis *Friedman ANOVA
* *Dependent *Mean responses of survey by topic *Independent *Specialist (Elementary physical education) *Non-specialist Analysis Paired-samples t-test
* *Dependent *Curriculum Base measurement of 60 key words *Independent *Intervention with use of flash cards Analysis One-way ANOVA (within subjects)
* Independent groups
* There is no significant difference between bullying incidents in schools that provide training to parents on student behaviors and schools that do not Dependent Bullying *Independent *Parenting *Communicating *Volunteering *Decision-Making *Analysis *Chi -Square
* *Dependent *Likert scale (5 point) *Independent *Math teachers *Science Teachers *Social study teachers Analysis *Kruskal-Wallis
* *Dependent *SAT-9 Reading *Independent *Grade * 6 th, 7 th & 8 th *Year * 2006, 2007, 2008 *Analysis *3 X 3 Factorial ANOVA (between subjects)
* Related Groups
* *Dependent *Cognitive functioning *Independent *Aerobic exercise * Yes & No *Memory strategies * Yes & No *Analysis *Cochrans Q test
* *Dependent *Participation *Independent *Divisions *Sport *Analysis *Friedman Two way ANOVA
* *Dependent * *Attitude toward science *Gender *Ethnicity *Major *Analysis *Factorial ANOVA