# B07 Tutorials + Midterm prep

Unlock Document

University of Toronto Scarborough

Psychology

PSYB07H3

Douglas Bors

Fall

Description

Tutorial 1
October 25, 2013 11:48 PM
Variable - A property of an event or an object that can take on diff values.
-Ex: height, weight, eye colour, gender, temperature, pet type
Independent Variable - The one you manipulate
Dependant Variable - The result that you measure, dependent on the IV
Discrete Variable - Groups; Gender, Pet type
Continuous Variable - Temperature (Something you can’t break into groups)
Qualitative Type - Categorical Data; (Numbers do not matter) Gender
Quantitative Type - Measurement Data; Grades on a test
Scales
Nominal - doesn’t differ along any dimension
-categorical
-cat vs dog, female vs male
-(defined by their name, numbers don’t matter)
Ordinal - order along a continuum
-nothing is implied about the numerical differences on scale
Interval - numerical differences between the scale points are meaningful
-e.g Temperature
-doesn’t imply a true zero, therefore you cannot do ratios
Ratio - same as interval but it has a true zero
-e.g Weight
Measure of central tendency
-A set of measure that reflect where on the scale the distribution(A set of values,
variable) is centered
-Mean (Average) - Sum of all values / num of values
-Median - Point at 50% above and below
-location = (N + 1) / 2
-Mode - most common score
3 5 2 5 1 4 2 2
-Mean = Sum/Num of vals = 24/8 = 3
-Median - 1 2 2 2 3 4 5 5
- Location = n + 1 / 2 = 9 / 2 = 4.5
- Median is 2.5 (Between two numbers)
-Mode - 2 (most common number) Tutorial 2
Tuesday, September 24, 2013 10:14 AM
1. Researcher wants to look at voting behavior in a small town. There are 7000people who are
eligible to vote. The interviewer contacted5000 individuals, 70% of those contacted agreed to
interview.
What is the population? 7000 individuals who can vote
Who is the sample? 70% of the 5000 = 3500
2. For a negatively skewed distribution with a mode of 25, and a median of 20, the mean is
probably..
A. >25
B. <20
C. Between 20 and 25
The answer is (B) since the tail is skewed to the negative side.
3. 1 1 2 3 3 3 3 4 5 5
1. Make frequency table
Score Freq
1 2
2 1
3 4
4 1
5 2
2. Graph it, score on x-axis, freq on y-axis. Title it, label it, even intervals. Histogram, the
interval tick is the mean.
4. CompositeMean - a.k.a Grand mean
X_1 = 6.3 N_1 = 10
X_2 = 10 N_2 = 5
The bigger the N, the smaller the variance will be.
X_m = SumOf(X)/N
SumOf(X) = X_m * N
Grand mean = (63 + 50) / 15 = 7.5333
Sum of the scores, divided by the combined count
5. Measure of Dispersion
1. Variance
2. Range
3. Interquartile range
i. Get the median, then split and get the median for the first half and the second half.
The point from the resulting first half median to the second half median is the IQR
Given a distribution, you can tell what the measure of central tendency is, and how wide it is. Given a distribution, you can tell what the measure of central tendency is, and how wide it is.
S^2 = ∑(X - Y_ )^2 / N - 1 - For samples
O^2 = ∑(X - Y_)^2 / N - For populations
SD = sqrt(S^2) Tutorial 3
Tuesday, October 1, 2013 10:10 AM
Z-Scores
-Standardizing
-Compare raw scores from different distributions
Z = (x - u) / o_
Ex. Matt's height - Male
Mary's height - Female
Is Matt taller than Mary, relative to their gender group?
Matt - 170 cm Y_ = 170 cm S = 5.6 cm
Mary - 150 cm Y_ = 140 cm S = 2.3 cm
Ex 2. Assume u = 100, o_ = 15, z = ? For x = 19
(19 - 100) / 15 = -5.4
Ex 3. Assume u = 95, o_ = 20, x = ? For z = 2
(x - 95) / 20 = 2
40 + 95 = 135
x = 135
-If you standardize all scores in a distribution, then we would have a z-score distribution.
-The shape of this distribution would be the same as the original but u

More
Less
Related notes for PSYB07H3