Psych stats: Prelims and Midterms

Psychology and Behavior

- Behavior and Mental Processes: The
study of psychology involves understanding Types of Data
how humans think, feel, and behave. - Quantitative vs Qualitative Data:
Quantitative data refers to numerical
Science of Psychology information, while qualitative data
refers to descriptive data.
● Scientific Method

Empirical (or Empiricism): Refers to

knowledge that is gained through Data Collection and Sampling
observation and experimentation. Techniques
Characteristics
Data Collection Methods
1. Observable: Psychological
phenomena must be something that - Population: The total count of individuals
can be observed or seen. or elements that a study focuses on.

2. Measurable: These phenomena - Sample: A subset of the population that

should be quantifiable. is used to represent the whole.

3.Testable: They should allow for - Attributes: Specific characteristics or

hypotheses that can be tested through features that are studied.
experimentation.
- Traits: Distinguishable qualities or
Statistics in Psychology characteristics of behavior.

● Fourfold Process - Behavior and Mental Processes: A

focus on how people think and act in
- Data Collection: Gathering of relation to their environment.
information or data, which serves as the
foundation for any analysis. ● Sampling Techniques

- Data Organization: Presentation of - Probabilistic Sampling: Each member of

data in a structured form, such as charts or the population has a known, non-zero
graphs. chance of being selected.

- Data Analysis: Processing and - Non-Probabilistic Sampling: Selection

examining data to extract meaningful does not involve random sampling, leading
patterns or conclusions. to non-representative samples.

- Interpretation: Understanding the

results and forming conclusions based on
the analysis.
Psych stats: Prelims and Midterms

● Types of Data in Statistics

1. Raw Data: Data in its original, ● Inferential Analysis(Parametric)

unprocessed form.
- This involves statistical techniques that
2. Treated Data: Data that has been generalize findings from the sample to the
processed for analysis. larger population.

3. Descriptive Statistics: Focuses on Key Statistical Tools and Terms

summarizing data without making
conclusions beyond the data 1. Central Tendencies: Mean, median, and
(non-parametric). mode that represent the center point of a
dataset.
4. Inferential Statistics: Draws conclusions
from data that go beyond the immediate 2. Variability: Measures how much data
dataset, often using probabilistic techniques varies from the average.
(parametric).
Comparison test:

Chi-Square: A test used to determine

● Measurement and Testing in whether there is a significant association
Statistics between categorical variables.

Variables Correlation and Relationship:

- Measurement: Assigning quantifiable - Spearman-Rho: A non-parametric test

values to variables. that measures the strength and direction of
association between two ranked variables.
- Testing: Verifying if variables can be
statistically analyzed. Scales of Measurement (NOIR)

● Descriptive Analysis - Nominal, Ordinal, Interval, Ratio:

(Non-Parametric) Different types of scales used to categorize
and measure variables.
Slovin's Formula: A formula used to
determine the sample size from a - Nominal: The simplest measurement
population: scale we can use to label variables is a
𝑁 nominal scale. A scale used to label
Formula; 𝑛 = 2 variables that have no quantitative values:
1+𝑁𝑒
Where
They have no natural order. For example,
N= is the population size we can’t arrange eye colors in order of
e= is the margin of error. (e= alpha level ) worst to best or lowest to highest.
Psych stats: Prelims and Midterms

● Categories are mutually and the median tells us the “middle”

exclusive. For example, an value
individual can’t have both blue and
brown eyes. Similarly, an individual
can’t live both in the city and in a Interval: A scale used to label variables that
rural area. have a natural order and a quantifiable
● The only number we can calculate difference between values, but no “true
for these variables are counts. For zero” value.
example, we can count how many
individuals have blonde hair, how ● These variables have a natural
many have black hair, how many order.
have brown hair, etc. ● We can measure the mean,
● The only measure of central median, mode, and standard
tendency we can calculate for deviation of these variables.
these variables is the mode. The ● These variables have an exact
mode tells us which category had difference between values. Recall
the most counts. For example, we that ordinal variables have no exact
could find which eye color occurred difference between variables – we
most frequently. don’t know if the difference between
“very satisfied” and “satisfied” is the
Ordinal: The next type of measurement same as the difference between
scale that we can use to label variables is “satisfied” and “neutral.” For
an ordinal scale. A scale used to label variables on an interval scale,
variables that have a natural order, but no though, we know that the difference
quantifiable difference between values; between a credit score of 850 and
800 is the exact same as the
difference between 800 and 750.
● They have a natural order. For ● These variables have no “true
example, “very satisfied” is better zero” value. For example, it’s
than “satisfied,” which is better than impossible to have a credit score of
“neutral,” etc. zero. It’s also impossible to have an
● The difference between values SAT score of zero. And for
can’t be evaluated. For example, temperatures, it’s possible to have
we can’t exactly say that the negative values (e.g. -10° F) which
difference between “very satisfied means there isn’t a true zero value
and “satisfied” is the same as the that values can’t go below.
difference between “satisfied” and
“neutral.”
● The two measures of central Ratio: A scale used to label variables that
tendency we can calculate for have a natural order, a quantifiable
these variables are the mode and difference between values, and a “true zero”
the median. The mode tells us value:
which category had the most counts
Psych stats: Prelims and Midterms

● These variables have a natural - Sampling Techniques: Important because

order. psychologists rarely study an entire
● We can calculate the mean, population, instead relying on a
median, mode, standard representative sample to make conclusions.
deviation, and a variety of other
descriptive statistics for these - Slovin's Formula: A useful tool in
variables. determining the number of participants
● These variables have an exact needed in a study to ensure the results are
difference between values. statistically significant.
● These variables have a “true
zero” value. For example, length, - Descriptive vs Inferential Statistics:
weight, and height all have a Descriptive statistics summarize the data
minimum value (zero) that can’t be (e.g., averages), while inferential statistics
exceeded. It’s not possible for ratio help make broader conclusions about the
variables to take on negative values. population.
For this reason, the ratio between
values can be calculated. For - Errors in Hypothesis Testing:
example, someone who weighs 200 Understanding the risks of making errors
lbs. can be said to weigh two times (Type I and Type II) helps researchers avoid
as much as someone who weights incorrect conclusions.
100 lbs. Likewise someone who is 6
feet tall is 1.5 times taller than
someone who is 4 feet tall. 1. Organization of Data:
- Data can be presented in three different
formats: textual (descriptions), graphical
(visual aids like charts), and tabular (tables).
This allows for easier understanding and
● Hypothesis Testing comparison of data.

- Type I Error: Rejecting a true null 2. Raw Data:

hypothesis (false positive). - Raw data is considered useless unless it
is organized. It needs to be categorized or
- Type II Error: Failing to reject a false null grouped for analysis.
hypothesis (false negative).
Summarization: 3. Frequency Distribution Table (FDT):

- Scientific Method: A cornerstone of - This table organizes data into categories

psychology, requiring observable and to display how frequently each category
testable phenomena. This ensures that occurs.
psychology stays rooted in empirical - Categorical FDT: deals with
research. non-numerical data (like gender or age
groups).
Psych stats: Prelims and Midterms

- Ungrouped FDT; is used for data

without classes (individual values).

- Grouped FDT: involves interval or ratio Data Set:

data, organizing it into classes for better 2.45, 2.12, 2.50, 2.68, 1.77, 2.19, 3.64,
readability. 2.50, 3.74, 1.89, 1.94, 2.52

4. Grouped Data Components: Tasks

- Grouped data uses several statistical 1. Mean: The average of the data points, is
elements such as arrays, classes, intervals, calculated by adding all the data points and
and frequencies. Sturge's formula helps dividing by the number of data points.
estimate the number of classes for grouped
data. 2. Median: The middle value when the data
points are arranged in order, requires
5. Measures of Data: sorting the data and selecting the middle
- Central Tendencies: Measures like value (or the average of the two middle
mean, median, and mode are used to values if there is an even number of data
summarize data by providing a central point. points).
- Variability: Measures like range and
standard deviation assess how much the 3. Average Deviation: The average of the
data varies. absolute differences between each data
- Position: Percentiles, quartiles, and other point and the mean, helps to understand the
positional measures help identify the dispersion of the data from the mean.
relative position of data points within a
distribution. ( 4. Standard Deviation: A measure of how
spread out the values are around the mean.
6. Types of Analysis:
- Descriptive Analysis involves 5. Variance: The square of the standard
summarizing data using basic statistics. deviation, shows how much the data points
- Inferential Analysis goes beyond differ from the mean squared, providing an
describing data to making predictions or overall idea of data spread.
inferences about a larger population based
on sample data. Tests like T-tests, ANOVA,
and correlation are used to determine Mean (Average):
relationships or differences between
variables. The mean is the sum of all values divided
by the number of values.
This structure gives a comprehensive Mean =
overview of organizing and analyzing data 2.45+2.12+2.50+2.68+1.77+2.19+3.64+2.50+3.74+1.89+1.94+2.52
using various statistical tools, both 12
descriptive and inferential, for effective 29.94
Mean = 12 = 2. 50 ( from 2.495 )
decision-making.
Psych stats: Prelims and Midterms

Median - First, subtract the mean from each data

point, square the result, and then find the
The median is the middle value in a sorted average of these squared differences
list. If the list has an even number of values, (variance):
the median is the average of the two middle
values. (1.77 - 2.50)^2= 0.5329
(1.89 - 2.50)^2= 0.3721
Array: (1.94 - 2.50)^2= 0.3136
1.77, 1.89, 1.94, 2.12, 2.19, 2.45, 2.50, (2.12 - 2.50)^2= 0.1444
2.50, 2.52, 2.68, 3.64, 3.74 ( 2.19 - 2.50)^2= 0.0961
(2.45 - 2.50)^2= 0.0025
2.45+2.50 (2.50 - 2.50)^2= 0
Median = 2
= 2. 48 (from
(2.50- 2.50)^2= 0
2.475) ( 2.52 - 2.50)^2= 0.0004
(2.68 - 2.50)^2= 0.0324
Average Deviation: The average deviation (3.64- 2.50)^2= 1.2994
is the mean of the absolute differences (3.74 - 2.50)^2= 1.5376
between each data point and the mean.
4.3314/ 12-1= 3.3314/11 = 0.3937 (
- First, calculate the difference between variance )
each data point and the mean: Take the square root of the average of
these variances to find the standard
| 1.77 - 2.495|= 0.725 deviation:
|1.89 - 2.495|= 0.605 0. 3937 = 0.627
|1.94 - 2.495|= 0.555
|2.12 - 2.495|= 0.375
| 2.19 - 2.495|= 0.305 T-Test, Z-Test, Hypothesis Testing
|2.45 - 2.495|= 0.045
|2.50 - 2.495|= 0.005 ● Two Tailed Test
|2.50- 2.495|=0.005 - Ho: Parameter = Specific Value
| 2.52 - 2.495|= 0.025 - Ha: Parameter ≠ Specific Value
|2.68 - 2.495|= 0.185 ● Right Tailed Test
|3.64- 2.495|= 1.145 - Ho: Parameter = Specific Value
|3.74 - 2.495|= 1.245 - Ha: Parameter ＞Specific Value
● Left Tailed Test
0.725+0.605+ 0.555+0.375+0.305+0.045+0.005+0.005+ 0.025+0.
12 - Ho: Parameter = Specific Value
5.22
= 12 = 0.44 ( from 0.435 ) - Ha: Parameter ＜Specific Value

Standard Deviation:
Standard deviation measures the
dispersion of the data points from the mean.
Psych stats: Prelims and Midterms

Hypothesis Testing proportion to a known

proportion.
1. Ho ○ Two-sample z-test: Used
The null hypothesis is a statement that when comparing the
indicates no effect or no difference. It is the proportions of two
hypothesis that the researcher tries to independent samples.
disprove.

2. Ha 5. Compute
The alternative hypothesis is a statement
that indicates the presence of an effect or a One Sample T-Test:
difference. It contradicts the null hypothesis
(𝑥−µ)
3. Level of Significance 𝑡= 𝑛
The level of significance (α) is the 𝑠
probability of rejecting the null hypothesis Where:
when it is true. Common levels are 0.05 t= computed t
(5%), 0.01 (1%), or 0.10 (10%). 𝑥= sample mean
µ= population mean
4. Appropriate level of significance and Test n= sample Size
s= sample SD
T-test:

● Use a t-test when the sample size is

Two Sample Mean T- Test:
small (typically n<30n < 30n<30)
and/or the population standard
deviation is unknown. (𝑋−𝑌)−(µ𝑋−µ𝑌)
○ One-sample t-test: Used t=
when comparing the sample 2 2
(𝑛𝑥−1)𝑠𝑥 +(𝑛𝑦−1)𝑠𝑦 1 1
mean to a known value. ( 𝑛𝑥+𝑛𝑦−2
)( 𝑛𝑥
+ 𝑛𝑦
)
○ Two-sample t-test: Used
Where:
when comparing the means
of two independent samples. 𝑋= Mean of the 1st Sample
○ T-Test for two sample mean: 𝑌 = Mean of the 2nd Sample
df=nx+ny-2 µ𝑋= 1st population Mean
µ𝑌= 2nd population mean
Z-test:
2
𝑠𝑥= variance of the 1st sample
● Use a z-test when the sample size
2
is large (typically n≥30n \geq 𝑠𝑦= variance of the 2nd sample
30n≥30) and/or the population 𝑛𝑥= 1st sample size
standard deviation is known.
𝑛𝑦= 2nd sample size
○ One-sample z-test: Used
when comparing the sample
Psych stats: Prelims and Midterms

● If the test statistic exceeds the

critical value, reject Ho.
● If the test statistic does not exceed
Z-Test: the critical value, fail to reject Ho.

One Sample Z-Test:

(𝑥−µ)
𝑍= 𝑛
7. Interpretation
σ

● Reject Ho: There is sufficient

Two Sample Z-Test: evidence to support the alternative
hypothesis.
(𝑥−𝑌)−(µ𝑋−µ𝑌) ● Fail to reject Ho: There is
𝑍= insufficient evidence to support the
2
σ𝑥
2
σ𝑦 alternative hypothesis.
𝑛𝑥
+𝑛
𝑦

Where:
𝑋= Mean of the 1st Sample
𝑌 = Mean of the 2nd Sample SPSS
µ𝑋= 1st population Mean
µ𝑌= 2nd population mean Le Description Example SPSS
2 ve Variab
σ𝑥= Population SD of the 1st population l le
2
σ𝑦 = population SD of the 2nd population Type

N Categorical, Gender, Nomin

nx= sample size of the 1st population o no order Nationality, al
ny= sample size of the 2nd population mi Eye color
na
l

Or Categorical, Satisfaction Ordina

di with ratings, l
6. Decision na meaningful Education
l order level
Compare the test statistic to the critical
Int Numeric, Temperatur Scale
value (from the t or z table, depending on
er equal e in Celsius, (in
the test) for the given level of significance
va intervals, no IQ scores SPSS)
(α).
l true zero
Decision rule:
Psych stats: Prelims and Midterms

Ra Numeric, Height, Scale

tio equal Weight, (in
intervals, Age, SPSS)
true zero Income