Intro to Data Analysis 9/25/2003

Required Text: Reading Statistics and Research, Huck. 1 copy on 2-hour reserve in Cubberley Library.

Papers should be submitted via Email, as Word documents. They should be named using the following convention:

Lastnamefirstinitalpaper#.doc

(for example, my first paper will be vaughannpaper1.doc)

Goal: become critical consumers of quantitative research.

Do not eat during class (implied)

160 (Statistical Concepts in Education) is targeted toward doctoral students who are going on to take more statistics.

SPSS workshop – 401A in the course schedule Mondays 3:15 – 5:05.

Amita (amitac@stanford.edu) will be running regular review sessions.

Ann’s philosophy: The most learning occurs when you’re actively involved in your own learning. My style in interactive lecturing. The material sounds easy in the beginning and gets increasingly hard – I need you to KEEP UP. Cramming won’t work here. Concepts build on each other; if you get behind you’ll be really lost later on.

There won’t be a lot of assignments, but I’m expecting you to ask questions, keep up with the readings, raise your hand when you don’t understand, explain concepts to fellow students when called on to do so, share with fellow students when you know something.

Office hours have not yet been set.

Descriptive statistics – describe set of people or things

Demographic: certain qualities of people, background of population (SES, race, quality, gender)

Descriptive statistic: the number (how many, how much)

Correlation: relationship between variables.

Reliability and Validity: When we collect data, we measure (people) in some way (tests, surveys, weighing, observation) – it’s very hard to get an exact measurement. Measurement instruments are never 100% accurate. We have to worry about how accurate those measuring instruments are (Validity), and how consistent those measuring instruments are (Reliability). This is how we know what we are interpreting.

Inferential statistics: where the heart of a lot of statistics is. What are the methods/process that allow us to study a small group of students and then generalize out to a larger group of people. How can a sample be representative of a whole?

Significance Testing: looking at the differences between two groups of people that have been studied. (T-testing)

Analysis of Variance: Testing for the differences between two or more groups of people

Linear Regression: Methods we use to predict and explain

Article Critiques:

• What’s well done?
• What’s not well-done?
• What do the numbers mean?
• How do they support the hypothesis?

Read “Tools A” and “Tools B” on the web for a guideline on what to look for and how to write it up.

These papers take a fairly long time to write, and all three are set toward the end of the quarter, so be prepared to budget your time accordingly. Feedback will be via Word comment function and a clear rubric will be developed for the work.

No late stuff past the end of the quarter. Also, deadlines are firm except in extreme circumstances (talk to Ann or Amita)

Access code for EDUC 150X on Blackboard – “iamin”

All course materials are available to Blackboard site. Ann will not bring copies of lecture notes – if you want them in hard copy you need to print them out yourself and bring them.

Why do we do research? What is research?

• Gather info to share with others
• Prove a point
• Find out differences within a group
• Solve problems
• Find answers to questions using the scientific method.
• Systematic approach that will yield interpretable, valid conclusions.

Descriptive: what it’s about, what the situation is.

Correlational: Predict/understand/identify relationships

Causal: Experimental/Comparative (control variables in comparative studies to isolate the causative factors) In education hard to control all the variables in order to do this.

Research Articles:

1. Abstract
2. Introduction
3. Background

i. Reasons

ii. Lit review

1. Nature of literature cited

2. Researcher bias

3. Rationale/need for study

4. Theoretical framework

5. Link of framework to research question.

6. Sufficiency of info

7. Usefulness of review

iii. Theoretical framework

4. Statement of Purpose

i. Goal of the research

ii. Problem statement

5. Hypotheses

i. Educated guesses

ii. Possible outcomes

iii. Research Question/Hypothesis

1. Narrow down – what do you have access to?

2. What don’t we already know? (contribution to existing knowledge)

6. Methods
1. Participants

1. Sample

a. Who are the subjects

b. How were they obtained/selected

2. Materials

1. Measuring instruments

a. What are they?

b. Reliability/validity

2. Is there sufficient detail for reader to go out and REPLICATE the study based on information contained in article?

3. Are sample & sampling method adequately described?

4. Are there weaknesses in study design?

7. Procedures
1. How study was conducted
2. What subjects did/were exposed to
8. Results
1. Text
2. Tables
3. Figures

i. Charts

ii. Graphs

iii. Other pretty pictures

4. Were the appropriate statistical techniques used given the kind of data?
5. Is the presentation of results clear and accurate?
8. Statistical Analyses
9. Discussion
1. Layman’s summary – the non-technical interpretation of the results section
2. Comparison of results with hypotheses and answers (or not) to original question/purpose
3. Analysis – why did the results turn out the way they did?
4. What were limitations of study?

i. Sample size, accuracy

ii. Lack of data or information on possible variable

5. Where to now? Steps for further research (into (d)?)
10. Review
11. Conclusions

References

IN STATISTICS WE NEVER PROVE ANYTHING. Things are “statistically likely” but ONLY indicate empirical knowledge, not absolute truth.

Homework: read article. DO NOT UNDERSTAND. Read introduction. Come to class prepared to discuss, in light of Tools A & B.