WEEK 1 - PRELIMS WHAT IS STATISTICS?

-​ Statistics is a study that deals with

TYPES OF DATA the collection, organization,
analysis, interpretation, and
1.​ Categorical (Qualitative Data) - presentation of data to
Categorical data represents understand or solve problems. It
labels, groups, or categories. It also helps in making informed
does not havea numerical decisions based on numerical/
meaning but is used to classify empirical evidence.
observations. TYPES OF STATISTICS:
a.​ Nominal: Categories with 1.​ Descriptive Statistics –
no specific order (e.g., Summarizes and presents given
gender, colors, car data in a meaningful way which
brands). can be a representation of the
b.​ Ordinal: Categories with a entire population or a sample.
meaningful order/rank but ○​ Measures of Central
unequal intervals (e.g., Tendency: Mean, Median,
customer satisfaction Mode
ratings: Poor, Fair, Good, ○​ Measures of Dispersion:
Excellent). Range, Variance,
2.​ Scale (Quantitative Data) - Scale Standard Deviation
data consists of numerical values 2.​ Inferential Statistics – analyzes
that represent measurable sample data to make predictions
quantities like height, weight, or or inferences about a larger
temperature. population. It involves drawing a
a.​ Discrete: integers, whole conclusion.
numbers, or counting ○​ Hypothesis Testing
numbers (number of ○​ Confidence Intervals
employees, number of ○​ Regression Analysis
products sold, allowance).
b.​ Continuous: contains
fractions and decimals (
(e.g., weight, height,
temperature)

​ SCALE MEASUREMENT:
●​ Ratio - there is real zero.
●​ Interval - no real zero.
●​ Ordinal - ranking
●​ Nominal - numbers
 WEEK 2 - PRELIMS Measures of Variation:
Measures of Central Tendency: -​ refers to how spread out or
-​ It indicates the center of the dispersed the data points are in a
data set. Subject to outliers. dataset from the central
(vulnerability of data using mean tendency. It helps measure the
to compare). consistency or inconsistency of
1.​ Mean (Average) values and provides insight into
●​ The sum of all values divided by the reliability of data.
the number of values.
●​ Formula: Mean=∑ X/N 1.​ Range
●​ Best for: Normally distributed data ●​ The difference between the
without extreme values. highest and lowest values in a
2.​ Median dataset.(higher range means the
●​ The middle value when data is data are spreadout.)
arranged in ascending order. ●​ Formula:
●​ Best for: Skewed data or data Range=Maximum
with outliers. Value−Minimum Value
●​ More stable than descriptive ●​ Example: If the highest salary in a
because it is not susceptible to company is ₱80,000 and the
outliers. (like dili ma change if nay lowest is ₱20,000, the range is
dako nga amount.) ₱60,000.
●​ Limitation: Only considers two
PARAMETER - population values and ignores distribution.
STATISTICS(n) - SAMPLE 2.​ Variance:
●​ Measures the average squared
3.​ Mode difference from the mean.
●​ The most frequently occurring ●​ High variance = More spread out
value in a dataset. data.
●​ A dataset can have one mode ●​ Low variance = Data points are
(unimodal), two modes close to the mean.
(bimodal), or more (multimodal).
●​ Example: In 2, 3, 3, 4, 5, 5, 5, 6, the
mode is 5 (appears most
frequently).
●​ Best for: Categorical data or
identifying common values in a
dataset.
Standard Deviation Example Calculation:
-​ The square root of variance; Consider the dataset: 5, 7, 9, 12, 15, 18,
shows the average distance of 22, 26, 30
data points from the mean. ●​ Q1 (25th percentile) = 9
●​ Q3 (75th percentile) = 22
●​ IQR = 22 - 9 = 13
Formulas:

Interquartile Range (IQR)

Answer:
-​ The Interquartile Range (IQR) is a
measure of statistical dispersion
that shows the spread of the
middle 50% of a dataset/ divided
intro 4 equal parts(25%). It is useful
for identifying variability while
reducing the impact of outliers.
IQR 5 VALUES SUMMARY:
1.​ Maximum - X largest
2.​ Minimum - X smallest
3.​ Q3
4.​ Q4
5.​ Q1.
How to Calculate IQR:
1.​ Arrange the data in ascending
order.
2.​ Find the first quartile (Q1) – the
median of the lower half (25th
percentile).
3.​ Find the third quartile (Q3) – the NOTE: If 5 ra ang sample (n),
median of the upper half (75th automatically apply the IQR.
percentile). = Xsmallest, Q1,Q2,Q3,Xlargest
4.​ Compute the IQR using the
formula: IQR=Q3−Q1
N(big N)usually refers to a population
size, while n(small n) refers to a sample
size.
CO-VARIANCE VS. CORRELATION ○​ +1: Perfect positive
correlation
Covariance ○​ 0: No correlation
-Covariance measures the direction of The strength of the relationship can be
the relationship between two variables. categorized as:
It indicates how changes in one ●​ Weak: A correlation coefficient
variable are related to changes in close to 0
another. ●​ Moderate: A correlation
●​ Positive Covariance: When both coefficient between 0.3 and 0.7
variables move in the same ●​ Strong: A correlation coefficient
direction, meaning if one between 0.7 and 1
increases, the other increases as ●​ Very Strong: A correlation
well, and if one decreases, the coefficient close to 1
other decreases.
●​ Negative Covariance: When the
variables move in opposite
directions, meaning if one
increases, the other decreases,
and vice versa.
●​ Zero Covariance: A covariance
of 0 means there is no relationship
between the variables, indicating
that they are independent of
each other.

Correlation
- Correlation measures both the strength
and direction of the relationship
between two variables. Unlike
covariance, which only indicates the
direction of the relationship, correlation
provides a clearer picture of how
strongly the two variables are related.

Strength: It indicates the strength

of the relationship. The strength is
measured by the correlation
coefficient, which ranges from -1
to +1:
○​ -1: Perfect negative
correlation
 WEEK 3 - PRELIMS ●​ Moderate skew: Between -1 and
1.
Skewness ●​ Extreme skew: Below -1
- Skewness indicates the degree of (negatively skewed) or above 1
asymmetry or lack of symmetry in a (positively skewed).
data distribution. Can We Run Statistical Tests?
●​ Positively Skewed (Right ●​ Yes, you can run tests on skewed
Skewed): data, but you might need to
○​ The data is concentrated normalize or transform the data
on the left side of the (e.g., log transformation) to meet
distribution. assumptions of normality.
○​ The mean is greater than Kurtosis
the median, and the Kurtosis measures the "tailedness" of the
median is greater than the distribution and the extent of outliers.
mode. ●​ Leptokurtic:
○​ Example: Sales data, ○​ Steeper curve
where a few very high (light-tailed).
values pull the mean to ○​ High kurtosis (> 3).
the right. ●​ Platykurtic:
●​ Negatively Skewed (Left ○​ Flatter curve
Skewed): (heavy-tailed).
○​ The data is concentrated ○​ Low kurtosis (< 3).
on the right side of the ●​ Kurtosis values:
distribution. ○​ 3: Normal distribution
○​ The mean is less than the (mesokurtic).
median, and the median is ○​ Less than 3: Negative
less than the mode. kurtosis (platykurtic).
○​ Example: Retirement data, ○​ Greater than 3: Positive
where fewer people retire kurtosis (leptokurtic).
early, so the mean is lower. ●​ Kurtosis Classification:
Pearson’s Second Coefficient of ○​ Between -1 and +1:
Skewness Excellent.
This measures the skewness of a ○​ Between -2 and +2:
distribution: Acceptable.
●​ Symmetrical distribution: Between Measures of Variation
-0.5 and 0.5 (close to 0). Variation or dispersion measures how
●​ Negatively skewed: Between -1 spread out the data is. Key measures
and -0.5. include:
●​ Positively skewed: Between 0.5 ●​ Range: Difference between the
and 1. maximum and minimum values.
 ●​ Variance: Measure of the Correlation
average squared deviations from Correlation measures both the strength
the mean. and direction of the relationship
●​ Standard Deviation: Square root between two variables:
of the variance. ●​ Range: From -1 to +1.
●​ Interquartile Range (IQR): The ○​ +1: Perfect positive
difference between the 3rd correlation.
quartile (Q3) and the 1st quartile ○​ -1: Perfect negative
(Q1). correlation.
○​ IQR = Q3 - Q1. ○​ 0: No correlation.
○​ The median is also known ○​ Closer to +1: Stronger
as Q2. positive correlation.
○​ IQR is resistant to outliers. ○​ Closer to -1: Stronger
Variables negative correlation.
●​ Independent Variable (X): Cohen’s D
○​ Also called the predictor Cohen’s D measures the effect size or
or explanatory variable. the difference between two groups.
●​ Dependent Variable (Y): ●​ Weak: Small effect size.
○​ Also called the outcome ●​ Moderate: Medium effect size.
or result variable. ●​ Strong: Large effect size.
Relationship Between Two Numerical ●​ Very Strong: Very large effect
Variables size.
●​ Discrete vs. Discrete: Both
variables are discrete
(countable).
●​ Continuous vs. Continuous: Both
variables are continuous
(measurable).
Covariance
Covariance measures the direction of
the relationship between two variables:
●​ Positive Covariance: If the two
variables move in the same
direction.
○​ Example: High IQ tends to
correlate with high
academic performance.
●​ Negative Covariance: If the two
variables move in opposite
directions.
 WEEK 4 - PRELIMS Compares two
different groups
Statistical Analysis and Chi-Square Test (e.g., male vs.
Understanding Frequency and female
Percentages intelligence).
1. Frequency – The number of times a ■​ Paired Sample
response appears in the dataset.​ T-test: Compares
Example: Counting how many the same group at
answered 4 or 5 in a survey. two different times
(e.g., pre-test vs.
2. Steps to Calculate Frequency in post-test).
excel: ○​ Regression Analysis & SEM
(Structural Equation
Modeling):
■​ More advanced
tests requiring
statistical software
like SPSS or AMOS.
Non-Parametric Tests
●​ Used when working with
categorical data (nominal or
ordinal).
●​ Measures association or
3. Determining the Range: relationships rather than
Use the MAX function to find the highest magnitude.
value in the dataset. ●​ Common Non-Parametric Test:
○​ Chi-Square Test
Parametric vs. Non-Parametric Tests ■​ Test of
Parametric Tests Independence:
●​ Used for numerical data when Examines the
making inferences about a relationship
population. between two
●​ Involves measuring magnitude, categorical
effect, or influence between variables.
variables. ■​ Test of Correlation:
●​ Common Parametric Tests: Examines the
○​ T-test: Determines if there is relationship
a significant difference between two
between two groups. numerical variables.
■​ Independent ○​ Example: Analyzing the
Sample T-test: relationship between
 gender and pet ○​ Fe values are computed
preference (dog vs. cat). for each cell using the
formula:
Chi-Square Test (χ² Test)
1.​ Understanding the 2x2 Table
Structure:
○​ Rows represent one
categorical variable (e.g.,
gender). Steps to Perform Chi-Square Test in Excel
○​ Columns represent 1.​ Input the observed frequencies
another categorical (Fo) into an Excel table.
variable (e.g., pet 2.​ Calculate the expected
preference). frequencies (Fe):
○​ Degrees of Freedom (df) Use the formula:​
Formula: =Frequency(Row Total * Column Total) /
df=(rows−1)×(columns−1) Grand Total) -itype ang number of
Key Terms: grand total.
○​ Fo (Observed Frequency):
The actual count from the 3.​ Compute the Chi-Square value
dataset. (χ²) using the formula:
○​ Fe (Expected Frequency): ○​ =CHITEST(Highlight
The theoretically expected Fo,Highlight Fe)
count, calculated as: 4.​ Find the Chi-Square critical value
Fe=N/k using the CHIINV function:
where N is the total sample =CHIINV(probability,degrees_of_f
size, and k is the number reedom)
of categories. ○​ Use a significance level
(alpha) of 0.05.
2.​ Example Data:​ ○​ Compare the computed
χ² value to the critical
Gender Dog Cat Total value:
○​ If χ² > critical value, reject
Male 5 6 11 the null hypothesis
(variables are
Female 20 13 33 dependent).
Total 25 19 44 ●​ If χ² ≤ critical value, fail to reject
the null hypothesis (variables are
○​ Fo values are directly independent).
taken from the table.
Conclusion H₀, we do not reject it, though this
●​ If χ² test shows a significant does not confirm that H₀ is true.
relationship, the two categorical 4. Level of Significance (Alpha, α)
variables are dependent. The level of significance (α) represents
●​ If not, there is no strong evidence the probability of rejecting a true null
of a relationship. hypothesis (Type I error). Common
values are 0.05 (5%) or 0.01 (1%),
1. Statistical Inference meaning there is a 5% or 1% risk of
Statistical inference is the process of concluding a false effect when none
using sample data to make conclusions exists.
about a population. It involves 5. One-Tailed Test vs. Two-Tailed Test
estimating population parameters (e.g., ●​ One-tailed test: Tests if a
mean, proportion) and testing parameter is greater than or less
hypotheses based on sample statistics. than a certain value (directional).
Methods of inference include estimation Example: Testing if a new drug
(confidence intervals) and hypothesis increases effectiveness.
testing. ●​ Two-tailed test: Tests if a
2. Tests of Hypothesis parameter is different from a
A test of hypothesis is a statistical certain value, regardless of
procedure used to decide whether a direction (non-directional).
statement (hypothesis) about a Example: Checking if a new
population parameter is supported by teaching method results in a
sample data. It involves formulating a different average test score.
null hypothesis (H₀) and an alternative 6. Independent Variable and
hypothesis (H₁), selecting a significance Dependent Variable in a Schema
level (α), computing a test statistic, and A schema representing the relationship
making a decision based on critical between independent and dependent
values or p-values. variables typically looks like this:
3. Statistical Hypothesis - Rejection vs. Independent Variable (IV) → Affects →
Non-Rejection of Hypothesis Dependent Variable (DV)
A statistical hypothesis is an assumption ●​ Independent Variable (IV): The
about a population parameter. The two variable that is manipulated or
possible decisions in hypothesis testing categorized (e.g., teaching
are: method).
●​ Rejection of the null hypothesis ●​ Dependent Variable (DV): The
(H₀): If the sample evidence outcome that is measured (e.g.,
strongly contradicts H₀, we reject student test scores).
it in favor of H₁. Example schema:
●​ Non-rejection of H₀: If the 📘 Teaching Method (IV) → Student
evidence is insufficient to reject Performance (DV)
7. Statement of the Problem 0.05), reject the null
A statement of the problem defines the hypothesis. Otherwise, fail
main research question(s) being to reject the null
investigated. It should be clear, concise, hypothesis.
and researchable, specifying the key 5.​ CHINV Function:
variables and the study's purpose. ○​ Purpose: Calculates the
Example:​ critical value of the
"Does implementing a digital budgeting chi-square distribution for a
system improve financial management given significance level
among small business owners?" and degrees of freedom.
Would you like further elaboration on ○​ The critical value obtained
any of these? from CHINV is compared
to the calculated
WEEK 5 - PRELIMS (LAST DISCUSSION) chi-square test statistic to
decide whether to reject
Excel Analysis Steps (Hospital length of or fail to reject the null
stay): hypothesis.

1.​ Identify Data Type: Hypothesis Testing Details:

○​ Compare ordinal vs.
ordinal data for the length 1.​ Research Question:
of stay and coverage ○​ Can you conclude that
categories. there is a relationship
2.​ Calculate Totals: between the length of
○​ Compute column totals hospital stay and
and row totals for hospitalization coverage
accurate data costs?
interpretation. 2.​ Hypotheses:
3.​ Compute Expected Frequency: ○​ Null Hypothesis (H₀):
○​ Necessary for statistical Length of stay and
analysis, possibly for a hospitalization costs are
chi-square test. not related.
4.​ CHITEST Function: ○​ Alternative Hypothesis (H₁):
○​ Purpose: Calculates the Length of stay and
p-value of the chi-square hospitalization costs are
test to determine the related.
significance of the 3.​ Significance Level (Alpha):
observed data compared ○​ α = 0.05 (threshold level of
to the expected data. significance)
○​ If the p-value is less than 4.​ P-value Interpretation:
the significance level (α =
 ○​ Less than 0.01: Highly ○​ If the difference is less than
significant. Thus, reject the 4: Assume equal
null hypothesis. variances.
○​ Greater than 0.05: Not
significant. Thus, do not T-test Analysis Steps(based on
reject the null hypothesis. assignment):
○​ Less than 0.05: Significant.
Thus, reject the null 1. Hypotheses Formulation:
hypothesis.
●​ Null Hypothesis (Ho):
5.​ Test of Hypothesis:
○​ Method B is as effective as
○​ Null Hypothesis (H₀): Reject
Method A.
or do not reject based on
p-value or chi-square
comparison.
○​ Alternative Hypothesis (H₁):
●​ Alternative Hypothesis (H1):
Accepted if H₀ is rejected.
○​ Method B is more effective
than Method A.
Conclusion and Interpretation:
○​ This suggests a One-tailed
●​ If the null hypothesis is rejected, it test since the direction of
indicates a significant effectiveness is specified.
relationship between the length
of hospital stay and the Type of T-test:
hospitalization cost coverage.
●​ One-tailed T-test:
●​ If the null hypothesis is not
○​ Used when the research
rejected, it suggests no significant
hypothesis suggests a
relationship between the two
specific direction (e.g.,
variables.
Method B is greater than
T-TEST ANALYSIS Method A).
Variance Assumptions: ○​ Rejection Region:
■​ If Right-tailed: The
●​ The decision on variance
rejection region is
assumption is based on the
on the right side of
difference between the
the distribution (for
variances:
hypotheses
○​ If the difference is greater suggesting an
than 4: Assume unequal increase or greater
variances. than).
■​ If Left-tailed: The
rejection region is
on the left side (for
 hypotheses ●​ For Positive T-stat or P-value:
suggesting a ○​ If T-stat > T-critical or
decrease or less P-value > Alpha → Reject
than). the Null Hypothesis
●​ Two-tailed T-test: ○​ If T-stat ≤ T-critical or
○​ Used when the hypothesis P-value ≤ Alpha → Do Not
does not suggest a Reject the Null Hypothesis
specific direction (e.g., ●​ For Negative T-stat or P-value:
there is a difference but ○​ If -T-stat < -T-critical or
unsure whether greater or P-value < Alpha → Reject
lesser). the Null Hypothesis
○​ Rejection Region: Both ○​ If -T-stat ≥ -T-critical or
ends of the distribution. P-value ≥ Alpha → Do Not
Reject the Null Hypothesis
2. Level of Significance (Alpha):
6. Conclusion
●​ It is the threshold for rejecting the
null hypothesis. 7.Insight
●​ Standard values are 0.05 or 0.01,
depending on the desired
confidence level

3. Calculate Test Statistic:

●​ T-stat (or T-value): The calculated

value from the T-test formula,
which measures the difference
relative to the variation in your
sample data.

4. Determine Critical Value:

●​ T-critical (or Tabulated T-value):

○​ Obtained from the
T-distribution table using
the degrees of freedom
and the chosen alpha
level.
○​ It serves as the cutoff point
for the rejection region.

5. Compare and Decision Rule: