Introductory Statistics Final Exam Review

Course Overview and Study Strategy

  • Foundational Reference: OpenStax Introductory Statistics 2e by Barbara Illowsky and Susan Dean (Textbook reference: https://openstax.org/details/books/introductory-statistics-2e).

  • Scope of Topics Covered:

    • Chapters 1–2: Sampling, data, and descriptive statistics.

    • Chapters 3–5: Probability and random variables.

    • Chapters 6–7: Normal distribution and the central limit theorem.

    • Chapters 8–9: Confidence intervals and hypothesis testing.

    • Chapters 10–13: Two samples, chi-square, regression, and ANOVA (as assigned).

    • Practice and Assessment: Calculator guide, 26-question mock final, answer key, and final checklist.

  • Strategic Priorities:

    • Prioritize Chapters 6–9 if study time is limited.

    • Practice selecting the correct formula, inputting parameters into a graphing calculator, and interpreting calculated results in verbal context.

    • Confirm the specific curriculum scope and chapters included on the course final examination.

Sampling, Data, and Measurement Levels (Chapter 1)

  • Fundamental Definitions and Statistical Terminology:

    • Population: The entire collection or group of individuals, objects, or items being studied.

    • Sample: A representative subset selected from the population to be analyzed.

    • Parameter: A numerical characteristic or summary measure calculated for an entire population.

    • Statistic: A numerical characteristic or summary measure calculated from sample data.

    • Variable: A characteristic or attribute that can take on different values for different individuals.

    • Data: The actual values, measurements, or observations collected for a variable.

    • Descriptive Statistics: Methods and procedures used to summarize, organize, and present data in an informative manner.

    • Inferential Statistics: Methods and procedures that use sample results to make decisions, predictions, or generalizations about a broader population.

    • Qualitative Data: Categorical measurements describing attributes or qualities (e.g., eye color).

    • Quantitative Discrete Data: Numerical values that arise from a counting process and take on distinct, separate points (countable values).

    • Quantitative Continuous Data: Numerical values resulting from measurements that can take on any real value within a given interval (measurable values).

  • Population vs. Sample Notation:

    • Mean: Population parameter is denoted by μ\mu; Sample statistic is denoted by xˉ\bar{x}.

    • Standard Deviation: Population parameter is denoted by σ\sigma; Sample statistic is denoted by ss.

    • Variance: Population parameter is denoted by σ2\sigma^2; Sample statistic is denoted by s2s^2.

    • Proportion: Population parameter is denoted by pp; Sample statistic is denoted by p^\hat{p}.

    • Size / Count: Population total size is denoted by NN; Sample total size is denoted by nn.

  • Levels of Measurement:

    • Nominal: Data consisting of qualitative categories, labels, or names with no implied ordering or hierarchy (e.g., categories).

    • Ordinal: Qualitative or quantitative categories that possess a meaningful natural order or ranking, but differences between data points cannot be measured or are mathematically meaningless.

    • Interval: Quantitative data with meaningful ordered differences between values, but lacking an inherent absolute (true) zero point (e.g., temperature in degrees Fahrenheit, ∘F^{\circ}\text{F}, where 0∘F0^{\circ}\text{F} does not represent the complete absence of heat).

    • Ratio: Quantitative data with meaningful intervals and an inherent absolute (true) zero point representing none of the measured quantity (e.g., weight, where zero indicates no weight, and ratios such as double weight are meaningful).

  • Sampling Methodologies:

    • Simple Random Sampling: Every possible sample of a given size nn has an equal probability of being selected from the population.

    • Systematic Sampling: Selecting a starting point at random and then choosing every kthk\text{th} individual from an ordered listing or stream.

    • Stratified Sampling: Dividing the population into distinct, non-overlapping subgroups (strata) based on specific characteristics, followed by drawing a random sample from every group.

    • Cluster Sampling: Dividing the population into natural clusters or groupings, randomly selecting entire clusters, and observing all individuals within the selected clusters.

    • Convenience Sampling: Selecting individuals or units that are easiest to access or reach, typically resulting in non-representative and biased data.

  • Frequency Distributions and Calculations:

    • Frequency: The absolute count of occurrences of a specific data value or within a given class.

    • Relative Frequency: The proportion or fraction of times a specific outcome occurs: Relative Frequency=FrequencyTotal Count\text{Relative Frequency} = \frac{\text{Frequency}}{\text{Total Count}}.

    • Cumulative Frequency: The running total of frequencies accumulating from the lowest value up through the current class.

    • Cumulative Relative Frequency: The running total of relative frequencies accumulating from the lowest value up through the current class.

    • Reference: OpenStax Chapter 1 Review (openstax.org/books/introductory-statistics-2e/pages/1-chapter-review).

Descriptive Statistics, Data Summaries, and Graphical Displays (Chapter 2)

  • Formulas and Measures of Center, Spread, and Position:

    • Sample Mean: Arithmetic average calculated by summing all observations and dividing by the total number of observations:     xˉ=∑xn\bar{x} = \frac{\sum x}{n}

    • Median: The physical middle observation in an ordered dataset arranged from lowest to highest. If nn is odd, it is the center value; if nn is even, it is the arithmetic average of the two central values.

    • Mode: The observation or value that occurs with the highest frequency in a dataset.

    • Midrange: The arithmetic midpoint between the extreme values: Midrange=min⁡+max⁡2\text{Midrange} = \frac{\min + \max}{2}.

    • Range: The total spread between the extreme values: Range=max⁡−min⁡\text{Range} = \max - \min.

    • Interquartile Range (IQR): The spread of the middle 50%50\% of the data, calculated as the difference between the third quartile (Q3Q_3) and the first quartile (Q1Q_1):     IQR=Q3−Q1\text{IQR} = Q_3 - Q_1

    • Sample Variance: The sum of squared deviations from the sample mean divided by degrees of freedom (n−1n - 1):     s2=∑(x−xˉ)2n−1s^2 = \frac{\sum (x - \bar{x})^2}{n - 1}

    • Sample Standard Deviation (ss): The positive square root of the sample variance:     s=s2s = \sqrt{s^2}

    • Population Standard Deviation (σ\sigma): The square root of the population variance, calculated dividing by the population size NN:     σ=∑(x−μ)2N\sigma = \sqrt{\frac{\sum (x - \mu)^2}{N}}

    • Grouped Mean: Weighted mean calculated from a frequency table using class frequencies (ff) and class midpoints (mm):     xˉ=∑fm∑f\bar{x} = \frac{\sum f m}{\sum f}

  • Distribution Skewness and Shape:

    • Right-Skewed (Positively Skewed): Data characterized by a prolonged tail extending toward larger values on the right. In general unimodal distributions: mode<median<mean\text{mode} < \text{median} < \text{mean}.

    • Left-Skewed (Negatively Skewed): Data characterized by a prolonged tail extending toward smaller values on the left. In general unimodal distributions: mean<median<mode\text{mean} < \text{median} < \text{mode}.

    • Symmetric Unimodal: A single-peaked, mirrored distribution where the mean, median, and mode coincide at the exact center.

    • Sensitivity of the Mean: The mean is non-resistant and is pulled in the direction of extreme values or the long tail, whereas the median is resistant to skewness and outliers.

  • Five-Number Summary, Quartiles, and Outlier Boundaries:

    • Five-Number Summary: A five-point descriptive set consisting of: min⁡\min, Q1Q_1 (first quartile or 25th percentile), Median\text{Median} (Q2Q_2 or 50th percentile), Q3Q_3 (third quartile or 75th percentile), and max⁡\max.

    • The Interquartile Range (IQR\text{IQR}) encompasses the middle 50%50\% of all ordered data values.

    • Outlier Identification Rules: A data value is identified as a potential outlier if it falls outside the outer fences:     Lower Outlier Boundary=Q1−1.5(IQR)\text{Lower Outlier Boundary} = Q_1 - 1.5(\text{IQR})     Upper Outlier Boundary=Q3+1.5(IQR)\text{Upper Outlier Boundary} = Q_3 + 1.5(\text{IQR})

    • Modified Box Plots: Graphical displays that truncate the whiskers at the smallest and largest non-outlier data points while plotting individual outliers as isolated asterisks or dots.

  • Graphical Displays of Data:

    • Histogram: A display for continuous numerical data featuring adjacent, touching bars where bar areas correspond to continuous class intervals.

    • Bar Graph: A display for qualitative or categorical data characterized by distinct, separated bars.

    • Frequency Polygon: A line chart constructed by connecting data points placed at class midpoints with heights equal to class frequencies.

    • Ogive: A line graph displaying cumulative frequencies or cumulative relative frequencies across upper class boundaries.

    • Box Plot: A visual display based on the five-number summary highlighting center, dispersion, and skewness.

    • Reference: OpenStax Chapter 2 Review (openstax.org/books/introductory-statistics-2e/pages/2-chapter-review).

Probability Foundations and Discrete Random Variables (Chapters 3–4)

  • Foundational Probability Principles:

    • Any probability P(A)P(A) satisfies: 0≤P(A)≤10 \le P(A) \le 1.

    • Sample Space (SS): The exhaustive set of all possible basic outcomes of a chance experiment.

    • Event: Any specified subset of outcomes from the sample space.

    • Independent Events: Two events AA and BB are independent if the occurrence of one does not affect the probability of the occurrence of the other.

    • Mutually Exclusive Events: Events that cannot occur simultaneously; their intersection is empty (P(A and B)=0P(A \text{ and } B) = 0).

  • Probability Rules and Formulas:

    • Classical / Equally Likely Outcomes Rule:     P(A)=Number of favorable outcomesTotal number of outcomes in sample spaceP(A) = \frac{\text{Number of favorable outcomes}}{\text{Total number of outcomes in sample space}}

    • Complement Rule:     P(not A)=1−P(A)P(\text{not } A) = 1 - P(A)

    • Addition Rule (OR Rule):     P(A or B)=P(A)+P(B)−P(A and B)P(A \text{ or } B) = P(A) + P(B) - P(A \text{ and } B)

    • Multiplication Rule for Independent Events (AND Rule):     P(A and B)=P(A)×P(B)P(A \text{ and } B) = P(A) \times P(B)

    • General Multiplication Rule for Dependent Events (AND Rule):     P(A and B)=P(A)×P(B∣A)P(A \text{ and } B) = P(A) \times P(B \mid A)

    • Conditional Probability Rule:     P(A∣B)=P(A and B)P(B)P(A \mid B) = \frac{P(A \text{ and } B)}{P(B)}

    • Concrete Addition Rule Example: Given P(A)=0.60P(A) = 0.60, P(B)=0.30P(B) = 0.30, and P(A and B)=0.20P(A \text{ and } B) = 0.20:     P(A or B)=0.60+0.30−0.20=0.70P(A \text{ or } B) = 0.60 + 0.30 - 0.20 = 0.70

  • Discrete Random Variables and Probability Distributions:

    • Expected Value (μ\mu): The weighted long-run mean of a discrete random variable XX:     μ=E(X)=∑xP(x)\mu = E(X) = \sum x P(x)

    • Binomial Distribution Criteria (Acronym BINS):

    • Binary: Outcomes on each trial are classified into two categories (Success or Failure).

    • Independent: Individual trials are statistically independent.

    • Number: There is a fixed, predetermined number of trials (nn).

    • Same Probability: The probability of success (pp) remains constant on each trial.

    • Binomial Formulas and Parameters:

    • Failure probability: q=1−pq = 1 - p

    • Mean: μ=np\mu = n p

    • Standard deviation: σ=npq\sigma = \sqrt{n p q}

    • Probability mass function for xx successes in nn trials:       P(X=x)=C(n,x)pxqn−x=(nx)px(1−p)n−xP(X = x) = C(n, x) p^x q^{n - x} = \binom{n}{x} p^x (1 - p)^{n - x}

    • Additional Discrete Distributions:

    • Geometric Distribution: Models the number of independent Bernoulli trials required to achieve the first success.

    • Hypergeometric Distribution: Models the probability of successes when sampling without replacement from a finite population composed of two distinct groups.

    • Poisson Distribution: Models the count of occurrences of an event over a specified interval of time, volume, area, or space.

  • Calculator Syntax for Binomial Distribution (TI-84 Functions):

    • Exact value (P(X=5)P(X = 5)): binompdf(n,p,5)

    • At most 5 successes (P(X≤5)P(X \le 5)): binomcdf(n,p,5)

    • Fewer than 5 successes (P(X<5)=P(X≤4)P(X < 5) = P(X \le 4)): binomcdf(n,p,4)

    • At least 5 successes (P(X≥5)=1−P(X≤4)P(X \ge 5) = 1 - P(X \le 4)): 1 - binomcdf(n,p,4)

    • More than 5 successes (P(X>5)=1−P(X≤5)P(X > 5) = 1 - P(X \le 5)): 1 - binomcdf(n,p,5)

Continuous Distributions, Normal Distributions, and the Central Limit Theorem (Chapters 5–7)

  • Continuous Probability Distributions:

    • The total area under any continuous probability density function (PDF) curve is strictly equal to 11.

    • The probability of a continuous random variable taking on any single exact point value is zero: P(X=c)=0P(X = c) = 0.

    • Continuous Uniform Distribution (X∼U(a,b)X \sim U(a, b)):

    • Probability between points cc and dd (a≤c<d≤ba \le c < d \le b):       P(c<X<d)=d−cb−aP(c < X < d) = \frac{d - c}{b - a}

    • Mean:       μ=a+b2\mu = \frac{a + b}{2}

    • Standard Deviation:       σ=b−a12\sigma = \frac{b - a}{\sqrt{12}}

    • Exponential Distribution (Waiting Times with Decay Parameter mm):

    • Right-tail probability: P(X>x)=e−mxP(X > x) = e^{-m x}

    • Left-tail (cumulative) probability: P(X≤x)=1−e−mxP(X \le x) = 1 - e^{-m x}

    • Mean and Standard Deviation:       μ=1m,σ=1m\mu = \frac{1}{m}, \quad \sigma = \frac{1}{m}

  • Normal Distribution Properties and Standardization:

    • Bell-shaped, perfectly symmetrical curve where the mean, median, and mode coincide.

    • The Empirical Rule (68-95-99.7 Rule for Bell-Shaped Data):

    • Approximately 68%68\% of observations lie within 1 standard deviation of the mean (μ±1σ\mu \pm 1\sigma).

    • Approximately 95%95\% of observations lie within 2 standard deviations of the mean (μ±2σ\mu \pm 2\sigma).

    • Approximately 99.7%99.7\% of observations lie within 3 standard deviations of the mean (μ±3σ\mu \pm 3\sigma).

    • Standard deviation scale positions: −3,−2,−1,Mean,1,2,3-3, -2, -1, \text{Mean}, 1, 2, 3.

    • Standard Score (zz-score) and Value Transformations:

    • Computing zz-score from raw data: z=x−μσ\quad z = \frac{x - \mu}{\sigma}

    • Computing raw value from zz-score: x=μ+zσ\quad x = \mu + z \sigma

    • A positive zz-score denotes a measurement situated above the mean; a negative zz-score denotes a measurement situated below the mean.

  • TI-84 Normal Distribution Calculator Operations:

    • Area below a value xx (P(X<x)P(X < x)): normalcdf(-1E99,x,μ,σ)

    • Area above a value xx (P(X>x)P(X > x)): normalcdf(x,1E99,μ,σ)

    • Area between values aa and bb (P(a<X<b)P(a < X < b)): normalcdf(a,b,μ,σ)

    • Finding a percentile value corresponding to a specified left cumulative area: invNorm(area,μ,σ)

  • Central Limit Theorem (CLT):

    • As the sample size nn increases, the sampling distribution of the sample mean xˉ\bar{x} approaches an approximately normal distribution, regardless of the underlying population shape, provided appropriate regularity conditions are satisfied.

    • Standard Error of the Mean (SE):     SE=σxˉ=σn\text{SE} = \sigma_{\bar{x}} = \frac{\sigma}{\sqrt{n}}

    • Standard Deviation of the Sample Sum (∑X\sum X):     σ∑X=σn\sigma_{\sum X} = \sigma \sqrt{n}

    • Standardized zz-Score for Sample Means:     z=xˉ−μσnz = \frac{\bar{x} - \mu}{\frac{\sigma}{\sqrt{n}}}

    • Standardized zz-Score for Sample Sums:     z=∑x−nμσnz = \frac{\sum x - n \mu}{\sigma \sqrt{n}}

    • Concrete CLT Application: For a population with μ=80\mu = 80, σ=12\sigma = 12, a sample size of n=36n = 36, and an observed sample mean of xˉ=84\bar{x} = 84:     SE=1236=126=2\text{SE} = \frac{12}{\sqrt{36}} = \frac{12}{6} = 2     z=84−802=2z = \frac{84 - 80}{2} = 2

Confidence Intervals and Sample Size Determination (Chapter 8)

  • General Structure of Confidence Intervals:   Confidence Interval=Point Estimate±Margin of Error\text{Confidence Interval} = \text{Point Estimate} \pm \text{Margin of Error}

  • Critical zz-Scores for Standard Confidence Levels:

    • 80%80\% Confidence Level: α=0.20\alpha = 0.20, α2=0.10\frac{\alpha}{2} = 0.10, Critical Value z∗=1.282z^* = 1.282

    • 90%90\% Confidence Level: α=0.10\alpha = 0.10, α2=0.05\frac{\alpha}{2} = 0.05, Critical Value z∗=1.645z^* = 1.645

    • 95%95\% Confidence Level: α=0.05\alpha = 0.05, α2=0.025\frac{\alpha}{2} = 0.025, Critical Value z∗=1.960z^* = 1.960

    • 98%98\% Confidence Level: α=0.02\alpha = 0.02, α2=0.01\frac{\alpha}{2} = 0.01, Critical Value z∗=2.326z^* = 2.326

    • 99%99\% Confidence Level: α=0.01\alpha = 0.01, α2=0.005\frac{\alpha}{2} = 0.005, Critical Value z∗=2.576z^* = 2.576

  • Selecting and Constructing Intervals:

    • Population Mean (μ\mu) with Population Standard Deviation (σ\sigma) Known:

    • TI-84 Function: ZInterval

    • Error Bound for a Population Mean (EBM):       EBM=z∗(σn)\text{EBM} = z^* \left(\frac{\sigma}{\sqrt{n}}\right)

    • Interval Formula: CI=xˉ±EBM\text{CI} = \bar{x} \pm \text{EBM}

    • Population Mean (μ\mu) with Population Standard Deviation (σ\sigma) Unknown:

    • TI-84 Function: TInterval

    • Student's tt-distribution is used with degrees of freedom: df=n−1df = n - 1

    • Error Bound for a Population Mean (EBM):       EBM=t∗(sn)\text{EBM} = t^* \left(\frac{s}{\sqrt{n}}\right)

    • Interval Formula: CI=xˉ±EBM\text{CI} = \bar{x} \pm \text{EBM}

    • Population Proportion (pp):

    • TI-84 Function: 1-PropZInt

    • Sample point estimate: p^=xn\hat{p} = \frac{x}{n}

    • Error Bound for a Proportion (EBP):       EBP=z∗p^(1−p^)n\text{EBP} = z^* \sqrt{\frac{\hat{p}(1 - \hat{p})}{n}}

    • Interval Formula: CI=p^±EBP\text{CI} = \hat{p} \pm \text{EBP}

  • Worked Confidence Interval Example:

    • Parameters: xˉ=72\bar{x} = 72, σ=10\sigma = 10, n=100n = 100, Confidence Level = 95%95\% (z∗=1.96z^* = 1.96).

    • Margin of error calculation:     EBM=1.96(10100)=1.96(1010)=1.96\text{EBM} = 1.96 \left(\frac{10}{\sqrt{100}}\right) = 1.96 \left(\frac{10}{10}\right) = 1.96

    • Interval endpoints: (72−1.96,72+1.96)=(70.04,73.96)(72 - 1.96, 72 + 1.96) = (70.04, 73.96).

    • Interpretation: We are 95%95\% confident that the true population mean falls between 70.0470.04 and 73.9673.96.

    • Meaning of Confidence Level: Describes the long-run success rate of the statistical estimation method (in repeated sampling, 95%95\% of constructed intervals will capture the true parameter).

  • Sample Size Determination Formulas:

    • Required Sample Size for Estimating a Mean (μ\mu):     n=(z∗)2σ2EBM2n = \frac{(z^*)^2 \sigma^2}{\text{EBM}^2}

    • Required Sample Size for Estimating a Proportion (pp):     n=(z∗)2p^(1−p^)EBP2n = \frac{(z^*)^2 \hat{p}(1 - \hat{p})}{\text{EBP}^2}

    • Practical Rules for Sample Size Computation:

    • Always round the resulting sample size calculation UP to the next whole integer.

    • If no prior estimate for p^\hat{p} is available, use p^=0.50\hat{p} = 0.50 to achieve a conservative maximum sample size.

  • Factors Influencing Interval Width:

    • Increasing the confidence level widens the interval.

    • Increasing the population or sample standard deviation widens the interval.

    • Increasing the sample size (nn) narrows the interval by reducing standard error.

    • Reference: OpenStax Chapter 8 Review (openstax.org/books/introductory-statistics-2e/pages/8-chapter-review).

Hypothesis Testing Principles, Decision Rules, and Error Types (Chapter 9)

  • Hypotheses Structure and Phrasing:

    • Null Hypothesis (H0H_0): Represents a statement of no effect, no difference, or status quo; always contains an equality sign (=,≤,≥=, \le, \ge).

    • Alternative Hypothesis (HaH_a or H1H_1): The research claim to be tested; contains strict inequality (≠,>,<\ne, >, <).

  • Tail Directions:

    • Left-tailed test: Ha:μ<μ0H_a: \mu < \mu_0 (values significantly less than the null value).

    • Right-tailed test: Ha:μ>μ0H_a: \mu > \mu_0 (values significantly greater than the null value).

    • Two-tailed test: Ha:μ≠μ0H_a: \mu \ne \mu_0 (values differing in either direction).

  • Standard Test Statistics:

    • Single Mean with Known σ\sigma (ZZ-Test):     z=xˉ−μ0σnz = \frac{\bar{x} - \mu_0}{\frac{\sigma}{\sqrt{n}}}

    • Single Mean with Unknown σ\sigma (tt-Test, df=n−1df = n - 1):     t=xˉ−μ0snt = \frac{\bar{x} - \mu_0}{\frac{s}{\sqrt{n}}}

    • Single Population Proportion (ZZ-Test):     z=p^−p0p0(1−p0)nz = \frac{\hat{p} - p_0}{\sqrt{\frac{p_0(1 - p_0)}{n}}}

  • Decisions and Error Analysis:

    • Definition of pp-value: The probability, assuming the null hypothesis H0H_0 is true, of obtaining a test statistic at least as extreme as the observed sample outcome in the direction specified by the alternative hypothesis HaH_a.

    • Significance Level (α\alpha): The allowable threshold probability of making a Type I error.

    • Decision Comparison:

    • If p≤αp \le \alpha: Reject H0H_0 (the result is statistically significant; possible error is Type I).

    • If p>αp > \alpha: Fail to reject H0H_0 (the result is not statistically significant; possible error is Type II).

    • At exact equality (p=αp = \alpha), follow instructor conventions.

    • Type I Error (α\alpha): Rejecting a null hypothesis H0H_0 when H0H_0 is actually true (a false alarm).

    • Type II Error (β\beta): Failing to reject a null hypothesis H0H_0 when H0H_0 is actually false (a missed detection).

    • Statistical Power: The probability of correctly rejecting a false null hypothesis: Power=1−β\text{Power} = 1 - \beta.

    • Logic of Hypothesis Testing: Failing to reject H0H_0 does not constitute mathematical proof that H0H_0 is true; it merely demonstrates an absence of sufficient evidence against it.

  • Standardized Conclusion Templates:

    • Original Claim is H0H_0 and Decision is Reject H0H_0: "There is sufficient evidence to warrant rejection of the original claim."

    • Original Claim is H0H_0 and Decision is Fail to Reject H0H_0: "There is not sufficient evidence to warrant rejection of the original claim."

    • Original Claim is HaH_a and Decision is Reject H0H_0: "The sample data support the original claim."

    • Original Claim is HaH_a and Decision is Fail to Reject H0H_0: "There is not sufficient sample evidence to support the original claim."

Hypothesis Testing Worked Example and Decision Practice (Chapter 9)

  • Step-by-Step Worked Right-Tailed Test:

    • Problem Context: Test the claim that the average exam score exceeds 7575.

    • Sample Data: xˉ=78\bar{x} = 78, σ=10\sigma = 10, n=100n = 100, α=0.05\alpha = 0.05.

    • Step 1: Formulate hypotheses: H0:μ=75H_0: \mu = 75, Ha:μ>75H_a: \mu > 75.

    • Step 2: Determine appropriate test: Right-tailed one-sample ZZ-test (population standard deviation σ\sigma is known).

    • Step 3: Compute test statistic:     z=78−7510100=31=3z = \frac{78 - 75}{\frac{10}{\sqrt{100}}} = \frac{3}{1} = 3

    • Step 4: Calculate the pp-value:     p=normalcdf(3,1E99,0,1)=0.00135p = \text{normalcdf}(3, 1\text{E}99, 0, 1) = 0.00135

    • Step 5: Formulate the statistical decision: Since p=0.00135<0.05p = 0.00135 < 0.05, reject the null hypothesis H0H_0.

    • Step 6: Interpret in context: There is sufficient evidence to support the claim that the average exam score exceeds 7575.

  • Section 9.4 Decision Practice Table:

    • Scenario 1: Given p=0.043p = 0.043, α=0.05\alpha = 0.05. Decision: Reject H0H_0. Support column: HaH_a. Possible Error: Type I.

    • Scenario 2: Given p=0.032p = 0.032, α=0.05\alpha = 0.05. Decision: Reject H0H_0. Support column: HaH_a. Possible Error: Type I.

    • Scenario 3: Given p=0.043p = 0.043, α=0.01\alpha = 0.01. Decision: Fail to reject H0H_0. Support column: H0H_0. Possible Error: Type II.

    • Scenario 4: Given p=0.0035p = 0.0035, α=0.01\alpha = 0.01. Decision: Reject H0H_0. Support column: HaH_a. Possible Error: Type I.

    • Scenario 5: Given p=0.15p = 0.15, α=0.10\alpha = 0.10. Decision: Fail to reject H0H_0. Support column: H0H_0. Possible Error: Type II.

    • Contextual Shorthand: An entry of H0H_0 in homework support shorthand signifies insufficient evidence to support HaH_a, not proof of H0H_0.

    • Validity Rule: Any reported pp-value falling outside the mathematical interval [0,1][0, 1] is invalid.

    • Reference: OpenStax Chapter 9 Review (openstax.org/books/introductory-statistics-2e/pages/9-chapter-review).

Advanced Inferential Topics: Two-Sample Tests, Chi-Square, Linear Regression, and ANOVA (Chapters 10–13)

  • Two-Sample Hypothesis Tests (Chapter 10):

    • Independent Means: Evaluated via a two-sample tt-test when population standard deviations are unknown; evaluated via a two-sample ZZ-test when population standard deviations are known.

    • Two-Sample Proportion Test: Two-proportion ZZ-test.

    • Dependent / Matched Pairs (e.g., before-and-after measurements on the same subjects): Evaluated via a paired tt-test on the single set of sample differences dd.

    • Independent Means Test Statistic (hypothesized difference equal to zero, unpooled variances):     t=xˉ1−xˉ2s12n1+s22n2t = \frac{\bar{x}_1 - \bar{x}_2}{\sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}}}

    • Paired Differences Test Statistic:     t=dˉ−μd,0sdnt = \frac{\bar{d} - \mu_{d,0}}{\frac{s_d}{\sqrt{n}}}

  • Chi-Square (χ2\chi^2) Distributions and Tests (Chapter 11):

    • Distribution Properties: Non-negative (χ2≥0\chi^2 \ge 0), asymmetric, and generally right-skewed; shape depends on degrees of freedom.

    • Four Primary Applications:

    • Goodness-of-Fit Test: Tests whether sample data conform to a hypothesized theoretical distribution.

    • Test of Independence: Tests whether an association or relationship exists between two categorical variables from a single population.

    • Test of Homogeneity: Tests whether different populations share the same distribution across categorical variables.

    • Test of a Single Variance: Tests whether a population variance σ2\sigma^2 equals a specified value.

    • Test Statistic for Goodness-of-Fit, Independence, and Homogeneity:     χ2=∑(O−E)2E\chi^2 = \sum \frac{(O - E)^2}{E}     where OO represents observed counts and EE represents expected counts.

    • Expected Frequencies for Two-Way Contingency Tables:     E=Row Total×Column TotalGrand TotalE = \frac{\text{Row Total} \times \text{Column Total}}{\text{Grand Total}}

    • Degrees of Freedom Formulas:

    • Goodness-of-Fit (with kk fixed probability categories): df=k−1df = k - 1

    • Test of Independence or Homogeneity (table with rr rows and cc columns): df=(r−1)(c−1)df = (r - 1)(c - 1)

    • Single Variance Test: df=n−1df = n - 1

  • Linear Regression and Correlation (Chapter 12):

    • Linear Correlation Coefficient (rr): Measures the strength and direction of a linear relationship between two quantitative variables; bounded by −1≤r≤+1-1 \le r \le +1.

    • Least-Squares Regression Line equation:     y^=a+bx\hat{y} = a + bx     where aa is the yy-intercept and bb is the slope.

    • Residual Calculation: The difference between the actual observed value and the predicted value:     Residual=y−y^\text{Residual} = y - \hat{y}

    • Coefficient of Determination (r2r^2): Represents the proportion of total variation in the dependent variable yy explained by the linear regression on the independent variable xx.

    • Concrete Regression Example: Given regression model y^=10+2x\hat{y} = 10 + 2x:

    • For an input value of x=5x = 5, the predicted value is y^=10+2(5)=20\hat{y} = 10 + 2(5) = 20.

    • If the observed actual value is y=23y = 23, the residual is: Residual=23−20=3\text{Residual} = 23 - 20 = 3.

    • Core Analytical Warnings:

    • Correlation does not imply causation.

    • Avoid extrapolation (making predictions outside the observed range of the explanatory variable xx).

  • One-Way Analysis of Variance / ANOVA (Chapter 13):

    • Hypotheses Structure:

    • H0:μ1=μ2=⋯=μkH_0: \mu_1 = \mu_2 = \dots = \mu_k (all population means are equal).

    • HaH_a: At least one population mean differs from the others.

    • Test Statistic (FF-ratio):     F=MSbetweenMSwithinF = \frac{\text{MS}_{\text{between}}}{\text{MS}_{\text{within}}}

    • Mathematical Assumptions:

    • Samples are independent random samples.

    • The populations under study are approximately normally distributed.

    • The populations possess equal variances (homogeneity of variances).

    • Post-Test Interpretation: Rejecting the null hypothesis in ANOVA indicates that not all group means are equal, but it does not specify which particular pairs or groups differ.

TI-84 Calculator Functions and Navigation Guide

  • Menu Access Commands:

    • Probability and Distribution Menus: Access by pressing 2nd then VARS ([DISTR]).

    • Statistical Tests and Confidence Intervals: Access by pressing STAT then selecting TESTS.

  • Comprehensive Function Directory:

    • Descriptive Statistics (mean, sum, sample SD, quartiles): 1-Var Stats

    • Exact Binomial Probability (P(X=x)P(X = x)): binompdf(n, p, x)

    • Cumulative Binomial Probability (P(X≤x)P(X \le x)): binomcdf(n, p, x)

    • Normal Cumulative Probability: normalcdf(lower, upper, μ, σ)

    • Inverse Normal (Percentile cutoff): invNorm(area, μ, σ)

    • Inverse Student's tt (Critical tt-value): invT(area, df)

    • Confidence Interval for Mean (σ\sigma known): ZInterval

    • Confidence Interval for Mean (σ\sigma unknown): TInterval

    • Confidence Interval for Single Proportion: 1-PropZInt

    • Hypothesis Test for Mean (σ\sigma known): Z-Test

    • Hypothesis Test for Mean (σ\sigma unknown): T-Test

    • Hypothesis Test for Single Proportion: 1-PropZTest

    • Two Independent Sample Means (tt-test): 2-SampTTest

    • Two Population Proportions (ZZ-test): 2-PropZTest

    • Chi-Square Test for Independence / Homogeneity: χ²-Test

    • Linear Regression Line Model: LinReg(a+bx)

    • Comparison of Three or More Means: ANOVA(

Practice Examination: Comprehensive Questions and Solutions

  • Questions 1–9:

    • Question 1: A total of 150 students are surveyed out of a total enrollment of 2,000. These 150 students represent what?

    • A. Population

    • B. Sample

    • C. Parameter

    • D. Variable

    • Correct Answer: B (A sample is a subset of the population).

    • Question 2: A subject's body weight represents what data type?

    • A. Qualitative

    • B. Discrete

    • C. Continuous

    • D. Ordinal

    • Correct Answer: C (Weight is continuous because it is a measurable continuous quantity).

    • Question 3: Selecting every tenth person entering a facility constitutes what sampling method?

    • A. Cluster

    • B. Stratified

    • C. Convenience

    • D. Systematic

    • Correct Answer: D (Systematic sampling involves selecting every kthk\text{th} individual after a random start).

    • Question 4: What is the arithmetic mean of the dataset 4, 6, 8, 10, 12?

    • A. 6

    • B. 7

    • C. 8

    • D. 10

    • Correct Answer: C (xˉ=4+6+8+10+125=405=8\bar{x} = \frac{4 + 6 + 8 + 10 + 12}{5} = \frac{40}{5} = 8).

    • Question 5: If Q1=12Q_1 = 12 and Q3=24Q_3 = 24, what is the Interquartile Range (IQR)?

    • A. 12

    • B. 18

    • C. 24

    • D. 36

    • Correct Answer: A (IQR=Q3−Q1=24−12=12\text{IQR} = Q_3 - Q_1 = 24 - 12 = 12).

    • Question 6: In a typically right-skewed distribution, which measure of central tendency has the largest value?

    • A. Mode

    • B. Median

    • C. Mean

    • D. All measures are equal

    • Correct Answer: C (Extreme observations in the prolonged right tail pull the mean above the median and mode).

    • Question 7: If the probability of event AA is P(A)=0.35P(A) = 0.35, what is P(not A)P(\text{not } A)?

    • A. 0.35

    • B. 0.65

    • C. 1.35

    • D. 0.50

    • Correct Answer: B (By the complement rule, P(not A)=1−0.35=0.65P(\text{not } A) = 1 - 0.35 = 0.65).

    • Question 8: If two events AA and BB are independent with probabilities 0.4 and 0.5, what is P(A and B)P(A \text{ and } B)?

    • A. 0.9

    • B. 0.1

    • C. 0.2

    • D. 0.45

    • Correct Answer: C (By independence, multiply probabilities: 0.4×0.5=0.20.4 \times 0.5 = 0.2).

    • Question 9: In a binomial distribution with parameters n=20n = 20 and p=0.3p = 0.3, what is the mean?

    • A. 3

    • B. 6

    • C. 7

    • D. 14

    • Correct Answer: B (μ=np=20×0.3=6\mu = n p = 20 \times 0.3 = 6).

  • Questions 10–18:

    • Question 10: Which calculator function is utilized to compute the probability of obtaining exactly four successes in a binomial experiment?

    • A. binomcdf

    • B. normalcdf

    • C. invNorm

    • D. binompdf

    • Correct Answer: D (binompdf computes point probabilities for exact counts).

    • Question 11: Given x=85x = 85, μ=75\mu = 75, and σ=5\sigma = 5, what is the standardized zz-score?

    • A. -2

    • B. 0.5

    • C. 2

    • D. 5

    • Correct Answer: C (z=85−755=105=2z = \frac{85 - 75}{5} = \frac{10}{5} = 2).

    • Question 12: According to the Empirical Rule, what percentage of normal data falls within two standard deviations of the mean?

    • A. 68%

    • B. 90%

    • C. 95%

    • D. 99.7%

    • Correct Answer: C (The 68-95-99.7 rule dictates approximately 95% falls within μ±2σ\mu \pm 2\sigma).

    • Question 13: Given a population standard deviation σ=20\sigma = 20 and sample size n=100n = 100, what is the standard error (SE)?

    • A. 0.2

    • B. 2

    • C. 10

    • D. 200

    • Correct Answer: B (SE=σn=20100=2010=2\text{SE} = \frac{\sigma}{\sqrt{n}} = \frac{20}{\sqrt{100}} = \frac{20}{10} = 2).

    • Question 14: What is the two-tailed critical value z∗z^* for a 95% confidence interval?

    • A. 1.282

    • B. 1.645

    • C. 1.960

    • D. 2.576

    • Correct Answer: C (Leaves an area of 0.025 in each tail).

    • Question 15: Which calculator test creates a confidence interval for a population mean when the population standard deviation is unknown?

    • A. ZInterval

    • B. TInterval

    • C. 1-PropZInt

    • D. 2-PropZInt

    • Correct Answer: B (The tt-distribution via TInterval is used when σ\sigma is unknown).

    • Question 16: If a confidence interval is calculated as (40,50)(40, 50), what is the margin of error?

    • A. 5

    • B. 10

    • C. 45

    • D. 90

    • Correct Answer: A (The margin of error equals half the total width: 50−402=5\frac{50 - 40}{2} = 5).

    • Question 17: When sample size nn is increased, what happens to the confidence interval width?

    • A. Wider

    • B. Narrower

    • C. Zero confidence

    • D. Unchanged

    • Correct Answer: B (Increasing nn reduces standard error, causing the interval to become narrower).

    • Question 18: If the research claim states that the population mean differs from 42, what is the alternative hypothesis HaH_a?

    • A. μ=42\mu = 42

    • B. μ>42\mu > 42

    • C. μ<42\mu < 42

    • D. μ≠42\mu \ne 42

  • Correct Answer: D ("Differs" translates to a two-tailed alternative hypothesis μ≠42\mu \ne 42).

  • Questions 19–26:

    • Question 19: An alternative hypothesis stating Ha:μ>30H_a: \mu > 30 requires which tail configuration?

    • A. Left

    • B. Right

    • C. Two

    • Correct Answer: B (A greater-than inequality specifies a right-tailed test).

    • Question 20: Given p=0.032p = 0.032 and α=0.05\alpha = 0.05, what is the correct statistical decision?

    • A. Reject H0H_0

    • B. Fail to reject H0H_0

    • C. Prove H0H_0 true

    • Correct Answer: A (Because p≤αp \le \alpha, reject H0H_0).

    • Question 21: Given p=0.07p = 0.07 and α=0.01\alpha = 0.01, what type of statistical error could possibly be made?

    • A. Type I

    • B. Type II

    • C. Neither

    • Correct Answer: B (Since p>αp > \alpha, the decision is fail to reject H0H_0; the possible error in this decision is Type II).

    • Question 22: If the original claim is HaH_a and the test decision is reject H0H_0, what is the appropriate verbal conclusion?

    • A. Reject claim

    • B. Support claim

    • C. Insufficient evidence

    • Correct Answer: B (Rejecting the null hypothesis provides sufficient sample evidence to support the original claim in HaH_a).

    • Question 23: Which statistical procedure is appropriate when comparing identical participants before and after a treatment?

    • A. Independent two-sample

    • B. Paired t

    • C. Chi-square GOF

    • D. One-proportion z

    • Correct Answer: B (Repeated measurements on the same participants constitute matched pairs analyzed via a paired tt-test).

    • Question 24: Which statistical test is used to evaluate an association between two categorical variables?

    • A. ANOVA

    • B. Paired t

    • C. Chi-square independence

    • D. One-sample z

    • Correct Answer: C (The chi-square test of independence assesses associations between categorical variables).

    • Question 25: A linear correlation coefficient of r=−0.92r = -0.92 indicates what type of relationship?

    • A. Strong positive

    • B. Weak negative

    • C. Strong negative

    • D. None

    • Correct Answer: C (Values close to −1-1 indicate a strong negative linear association).

    • Question 26: Which procedure is used to compare the population means of three independent groups?

    • A. ANOVA

    • B. One-proportion z

    • C. Binomial

    • D. Normalcdf

    • Correct Answer: A (One-Way Analysis of Variance compares three or more group means simultaneously).

Final Review Checklist and Seven-Day Study Schedule

  • Core Competency Checklist:

    • Distinguish population, sample, parameter, and statistic; classify data types and identify measurement levels.

    • Construct and interpret sampling techniques and frequency distributions.

    • Calculate mean, median, mode, variance, and standard deviation.

    • Compute quartiles, determine IQR, identify outliers, construct graphical displays, and identify skewness.

    • Compute AND, OR, complement, and conditional probabilities.

    • Solve binomial distributions and execute appropriate calculator functions.

    • Calculate standardized zz-scores and evaluate areas/percentiles via normalcdf and invNorm.

    • Apply the Central Limit Theorem and compute standard error.

    • Select correct confidence intervals, calculate margins of error, and determine necessary sample sizes.

    • Formulate null and alternative hypotheses and classify test tail directions.

    • Calculate test statistics, determine pp-values, make statistical decisions, and recognize error types.

    • Construct properly worded conclusions based on the original research claim.

    • Execute two-sample tests, chi-square tests, linear regression, and ANOVA models as assigned.

  • Seven-Day Structured Study Plan:

    • Day 1: Review Chapters 1–2 (Data, sampling, and descriptive statistics).

    • Day 2: Review Chapters 3–5 (Probability distributions and discrete random variables).

    • Day 3: Review Chapters 6–7 (Continuous uniform, exponential, normal distribution, and CLT).

    • Day 4: Review Chapter 8 (Confidence interval construction and sample size estimation).

    • Day 5: Review Chapter 9 (Hypothesis testing logic, formulas, error types, and conclusions).

    • Day 6: Review Chapters 10–13 (Two-sample inference, chi-square, regression, and ANOVA if included).

    • Day 7: Complete the 26-question practice exam, evaluate performance with the answer key, and review identified weaknesses.

  • Reference Source: OpenStax, Introductory Statistics 2e. Complete chapter reviews, end-of-chapter summaries, and formula reviews are accessible in the textbook table of contents.