Sie sind auf Seite 1von 15

So You Think You Know

Statistics!
Created by Ellie Leong – revised by Dr. Huselid

Hunter College of The City University of New York


Scales of Measurement
 Continuous or Scale (Ratio & Interval)
 Numerical scores (height, weight, IQ, test scores)

 Categorical
 Ordinal
 Ranked categories (size, year in school, placement in race)
 Nominal
 Categories without order (gender, hair color, career)
A survey asked 3 Questions:
(1) How many drinks on average do you have per drinking occasion?
(2) When did you have your first drink of alcohol?
(3) Where do you typically consume alcohol?

ID (1) Alcohol (2) Alcohol (3) Alcohol


Consumption Consumption Consumption
1 4 Before 14 Bar
2 3 14-18 Home
3 1 18-21 Restaurant
4 2 14-18 Party

What kind of data is each var?


Identify the Type of measures:
ID (1) Alcohol (2) Alcohol (3) Alcohol
Consumption Consumption Consumption
1 4 Before 14 Bar
2 3 14-18 Home
3 1 18-21 Restaurant
4 2 14-18 Party
5 1 After 21 Bar
6 5 Before 14 Home
7 4 After 21 Bar
8 3 14-18 Restaurant
9 2 18-21 Bar
Ratio Ordinal Nominal
Identifying Variables
 Independent Variable
 Controlled or changed (treatment condition, before/after)
 Sometimes just observed (babies whose mothers drank
during pregnancy, addiction, brain damage)

 Dependent Variable
 What is being measured (reaction time, scores on test,
rating)

 How many Levels? (for categorical vars)


 Gender – 2 levels (male, female)
 Hair color – 4 levels (brown, black, blond, red)
What is the scale of the IV? 6

Nominal or Ordinal Interval or Ratio


Choose the
Right Test
What is the scale of the DV? Correlation/
Flowchart
Regression

This may Interval or Ratio Nominal or Ordinal


be useful –
but doesn’t How many groups? Chi-Square
cover
repeated
ANOVA 1 2 >2
design

Is σ Are the samples


How many IVs?
known? independent (not matched)?

yes n 1 2
yes n
o
o

One- One- 1-Way 2-Way


sample z sample t ANOVA ANOVA

Independent Dependent
measures t measures t
For each round identify the following:

 Independent Variable
 Categorical or continuous
 # Levels (if categorical)

 Dependent Variable
 Categorical or continuous
 # Levels (if categorical)

 What statistic test should you use?


 One-Sample t-Test
 Independent Samples t-Test
 Related-Sample t-Test
Study 1
Dr. Rabbit wants to know if mood ratings of clinically
depressed patients increase after playing with animals.
He first gives each participant a test that has questions
with a Likert scale and records their individual scores.
Then, he exposes participants to 30-mins of play with a
pet of their choice. After play, he gives them the same
test again and records their scores.
What do we know?

 Independent: Interaction with animals


 Categorical
 2 levels – before, after (same group)

 Dependent: Mood ratings


 Continuous
Study 1
Do mood ratings increase after interactions with
animals?

Related Samples t-Test!


 One categorical IV, one group tested twice

 One continuous dependent variable


Study 2
College students nationwide get an average
of 7 hours of sleep. Do Hunter College
students significantly differ from the
national average?

What do we know?

 One variable: hours of sleep


 Continuous

 Population mean
Study 2
College students nationwide get an average of 7
hours of sleep. Do Hunter College students
significantly differ from the national average?

One-Sample t-Test!
 One continuous variable compared to a population mean
Study 3
Dr. E. Harmony wants to see if men and
women rate the secureness of their
long-term relationships differently. He
gives a survey with Likert scale ratings
of security to 20 men and 20 women.
What do we know?

 Independent: gender
 Categorical
 2 levels – male, female

 Dependent: secureness rating


 Continuous
Study 3
Do men and women rate security in long-term
relationship differently?

Independent Samples t-Test!


 One categorical IV with 2 separate groups

 One continuous DV
Study 4
Do students who work more than 20
hours a week have lower scores than
students who work less than 5 hours a
week?

What do we know?

 Quasi- Independent: hours employed each week


 Categorical
 2 levels – <5, 20+

 Dependent: exam scores or grades


 Continuous
Congrats! Choosing the right
Statistical Test is challenging!
 If you got all of them correct you are doing better than most
Graduate Students!

 If you got 75% correct you are doing extremely well.

 If you got 50% or less correct – you have learned some things
but keep working on it –

you can learn to decode word problems (ID the IV and DV)
and apply statistical tests appropriately!

Das könnte Ihnen auch gefallen