# The Process of Statistical Analysis in Psychology

# Web Activities

Click on the following links. Please note these will open in a new window.

*Activity 1:*

**Representation and Sample Size**

Your text discussed how larger samples are more likely to be representative of the population. In this activity, you will look at sample size and representation in a population.

Go to http://www.shodor.org/interactivate/activities/AdjustableSpinner/.

In the box labeled number of spins, put in 4. Then click Spin. Look at the box that says Probabilities (on the right). Compare the Experimental (what you got) to the Theoretical (what would be expected if equal proportions were obtained from each group). How do they compare?

Click New experiment. Change the number of spins to 20. Compare the Experimental to Theoretical. How do they compare?

Click New experiment. Change the number of spins to 50, then 100, then 150, clicking New experiment after each spin. What is the best option? How low do you think the sample size can be to get a good representation of the population?

*Activity 2:*

Your book talks about different types of sampling. In this activity, you will think about some of the different types of sampling.

Go to: http://www.myscienceproject.org/trick.html.

At the top they discuss the “Halloween candy hierarchy.”

- Identify the three strata.
- How do they attempt to account for the strata of Halloween candy?
- Is this actually stratified
*random*sampling? Generate an idea for how you could create a random stratified sample of Halloween candy.

*Activity 3:*

**Sample size and sampling error**

Your book discussed the discrepancy in scores between the sample and population (sampling error). In this activity, you will examine how sample size might impact a dependent variable.

Got to amazon.com and search for any item you like. Look for a best seller or one with at least 500 reviews, and another with fewer than 50 reviews. Also look for one with 5 stars and one with 3 stars as an average review.

For each item, hover over the arrow next to the average customer review stars and look at the distribution of stars. Compare the distribution of stars in each item.

- How many (number) low (<3) reviews do you see for the item with 500+ reviews? How many are there for the one with <50? How similar are these numbers? How did sample size impact the average rating for these items?
- How many (number) low reviews do you see for the item with an average 3-star rating? How many for the average 5-star rating? How did sample size impact the rating for these items?
- Which item would you be most confident in buying and why?