Scribd
Upload
9 views19 pages

Class02 Review and Measurement

Uploaded by

zouyikehappiest
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
9 views19 pages

Class02 Review and Measurement

Uploaded by

zouyikehappiest
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

Learning Objectives

• Remember key terms from research methods:


– Variable: Independent vs. dependent
– Data type: Samples vs. populations
– Statistics: Descriptive vs. inferential

• Describe the different types of variables:


– Four measurement scales
– Continuous vs. discrete

• Explain (very broadly) why we care about


variable types 1
Remember these terms?..
• Variable: Property that can have different values

• Independent Variable (IV):


Variable that is manipulated by the researcher

• Dependent Variable (DV):


Variable that is measured by the researcher

4
Remember these terms?..

Sample Population

Subset of all cases Complete set of cases

Described by sample statistics Described by parameters

Example: Sample mean Example: Population mean

5
Two Types of Statistics
• Descriptive statistics seek to understand
patterns in the sample

• Inferential statistics seek to infer whether


patterns generalize to the population
– Often inferential statistics quantify our confidence
– Will sample pattern show up in the population?

6
Describing political preferences:
“If the election were today, who would you vote for?”

7
“If the election were today, who would you vote for?”

• Sample: 447 residents of New Mexico who self-report they will vote
in the Democratic primary as of early January, 2020
– “N = 447”

• Population: All residents of New Mexico who self-report they will


vote in the Democratic primary

• Descriptive statistic: Sanders is top with 28%, Biden 2nd with 27%

• Inferential statistic: Maybe Sanders, maybe Biden… (+/- 4.6%)


– Population might reasonably be:
• Sanders 33%
• Sanders 23%

• Interpretation: “It’s a statistical tie”


– Sanders at the top in this sample, but we’re not confident this is true
in the population
8
Who’s going to win the election
tomorrow?
Number of Bernie
Poll Joseph Biden
voters polled Sanders

Maris 10 20% 80%

Onward 909 64% 36%

Digi 2,042 49% 51%

Johnson 40,000 50% 50%


9
10
Measurement Scales

Ratio

Interval

Ordinal

Nominal

11
Measurement scales
• Variables are quantified for statistical analyses
– Different measurement scales permit different
statistical techniques

1. Nominal scale: Values are arbitrary


– Also known as a “categorical variable”
– Example: undergraduate major
• Psychology = 1
• Engineering = 2
• Kinesiology = 3
– Nominal variables tell us only equivalence 12
Measurement scales
• Variables are quantified for statistical analyses
– Different measurement scales permit different
statistical techniques

2. Ordinal scale: Values are ranked


– But, interval between values may not be equal
– Polling example:
• 1st Sanders, 2nd Biden, 3rd Yang
• We know that Sanders had more support than Biden,
but we do not know how much more support
– Ordinal variables convey rank but not magnitude
13
Measurement scales
• Variables are quantified for statistical analyses
– Different measurement scales permit different
statistical techniques

3. Interval scale: Ranked w/equal intervals


– Example: Day of the month
• Today is the 7th, 14th will be exactly 168 hours from
now, 21st will be 336 hours from now…
• But there is no absolute zero point (no day ‘zero’)

14
Measurement scales
• Variables are quantified for statistical analyses
– Different measurement scales permit different
statistical techniques

4. Ratio scale: Ranked, equal intervals, & zero


– The value of 0 is possible and meaningful
– Example: Reaction time
• Zero means you responded simultaneously with the
presentation of the stimulus
• It took (literally) no time!
15
Measurement Scales

Calculate Ratios Ratio


Add & Subtract!!! Interval
Compare different things,
put them in order Ordinal
Count things that are alike Nominal

16
What measurement scales?
You want to know if dog weight
is related to the speed of their eating.

Weight is measured by Speed of eating is rated


putting dogs on a scale: by observers as:
Slow = 1
Medium = 2
Fast = 3

17
Discrete vs. Continuous
• Discrete variables have no possible values
between adjacent units on the scale
– Example: Number of dogs (1 or 2, but NOT 1.5!!)

• Continuous variables can have infinite


possible values between adjacent units
– Example: Weight of dog (1 gram, 2 g, or 1.5g)
• Or 1.5000000000000001 grams

18
Real Limits of Continuous Variables
• Continuous variables are always approximate
– Estimates depend on our instrument’s resolution
• Real limits are the values above and below the
recorded value (by ½ the smallest measuring unit)
– Measurement = 34.46lbs
– Smallest measuring unit = .01lbs (half = .005lbs)
– Real limits for this scale:
• Real lower limit = 34.455
• Real upper limit = 34.465

19
Significant Figures
“How many decimal places should I report?”
– Means, standard deviation
• Report 2 or 3 decimal places

– Correlation/regression coefficients, test-statistics


(like t, F, etc.)
• Always report 3 decimal places

– p-values (or ‘probability values’)


• Always report 3 decimal places
20
Rounding
Don’t worry about the text’s discussion of
rounding (pp. 38)

• Last digit equal to or greater than 5, round UP

• Last digit less than 5, round DOWN

21

You might also like