Show:
Experimental Design
Standard Experiments
Power
One Way ANOVA
100

What is the first step in the Experimental Checklist?

Define the Objectives of the Experiment

100

What is a completely randomized design?

Experimenter assigns experimental units to the treatment completely at random
100

What is considered our traditional cutoff for Power?

.80

100

What does the F statistic measure?

Total variability in the data

200

Why do we conduct a pilot test?

To practice the experiment and identify problems

200

What method is used to obtain the estimates for parameters in linear model?

Least Squares

200

What is Power?

Probability that we will identify a difference truly exists (reject the null)

200

What is the null hypothesis for ANOVA?

All of the treatment means are equal

300

When identifying the sources of variation? How is variance split?

Treatment factor and levels

Experimental units

Blocking factors, noise factors, and covariates

300

Why do we use randomization?

To prevent systematic and personal bias

300

What is the probability of a Type I error when conducting 5 t-tests?

.23

300

What is the F distribution shape based on?

both degrees of freedom (df1 and df2); df1 = treatment, df2 = sample

400

Professor Moon is conducting an experiment examining the effects of studying (time spent: 30 mins, 60 mins, 90mins) on test scores. What are some anticipated difficulties that may arise?

Study time doesn't account for quality of studying

Not controlled environment so students may be distracted

The exam may not be accurate measurement of the material

400

What is a key principle for minimizing influence of outside variables?

Random assignment

400
Professor Moon conducted an experiment and didn't find significance. He collected data from 50 participants from the undergraduate psychology pool. He had his alpha set to .05 and randomized conditions. He had 5 experimental conditions. What should he do next?

Conduct a post-hoc power analysis to determine what his power of experiment is

400
What kind of function are we expecting with ANOVA?

Linear model

500

Professor Moon is analyzing the data from his experiment on training and performance. He has an independent variable (online, hybrid, in-person) and a dependent variable (performance). He conducted multiple t-tests between each of the conditions and noticed that those who had in-person training performed significantly higher than the online on the exam. What’s wrong with this scenario?

Type I error is increased because he conducted multiple t-tests

500

Professor Moon is using a Block design where he divided experimental units into blocks. What is his associated model?

Response = constant + effect of block + effect of treatment + error

500

What does power depend on?

Design

Significance level (or alpha)

Parameter (effect size)

Sample size

500

What is the formula for completely randomized design?

Response = constant + effect of treatment + error