Data Collection Methods and Statistical Sampling Principles

Four Steps of the Statistical Problem-Solving Process

• Clarify the research problem and formulate one or more valid statistical questions.

• Plan and carry out an appropriate plan to collect the data.

• Apply appropriate graphical and numerical methods to analyze the data.

• Interpret the results and make conclusions based on the data analysis.

Major League Baseball Roster Data and Population Parameters

• Opening day active roster allowance: $26$ players per Major League Baseball ($\text{MLB}$) team.

• Total number of teams in $\text{MLB}$: $30$ teams.

• Total population size of players across all opening day rosters: $N = 30 \times 26 = 780$ players.

Probability Sampling Designs and Implementation Procedures

• Simple Random Sample ($\text{SRS}$) of $100$ Players:

  • Labeling step: Assign a unique numerical label to each player from $1$ to $780$ ($1-780$).

  • Randomization step: Use a Random Number Generator ($\text{RNG}$) to generate $100$ unique random integers between $1$ and $780$ without replacement.

  • Selection step: Identify and select the $100$ players corresponding to the generated numbers and survey them.

• Stratified Random Sample of $90$ Players:

  • Stratification and labeling step: Group players naturally by team to serve as $30$ distinct strata. Label the $30$ teams, and label the players within each individual team from $1$ to $26$ ($1-26$).

  • Randomization step: Using an $\text{RNG}$, generate $30$ distinct sets of $3$ unique integers from $1$ to $26$, one set for each team.

  • Selection step: Select the $3$ identified players from each of the $30$ teams to survey, yielding a total stratified sample of $30 \times 3 = 90$ players.

• Cluster Sample of $78$ Players:

  • Cluster identification and labeling step: Define each team as a cluster containing $26$ players. Label the teams from $1$ to $30$ ($1-30$).

  • Randomization step: Use an $\text{RNG}$ to select whole clusters. Since each team contains $26$ players, select $3$ clusters ($\frac{78}{26} = 3$ teams; noted in field responses as generating a set from $1-30$).

  • Selection step: Include and survey all $26$ players rostered on each of the selected teams, producing a total cluster sample of $3 \times 26 = 78$ players.

• Systematic Random Sample of $78$ Players:

  • Labeling step: Assign a unique sequential integer from $1$ to $780$ ($1-780$) to all players.

  • Interval determination and randomization step: Compute the sampling interval $k$ using the ratio of total population to sample size:     $k = \frac{780}{78} = 10$     Use an $\text{RNG}$ to select a single random integer from $1$ to $k$ (such as a random start at $1\text{st } \# = 6$ selected from $1-10$ or within an initial range up to $30$).

  • Selection step: Starting at the randomly selected starting position, select every $k\text{th}$ (e.g., $10\text{th}$) player systematically down the roster list until exactly $78$ players are selected for the survey.

Freshmen Orientation Cluster Sampling Analysis

• Scenario and Institutional Context:

  • Setting: Freshmen orientation day hosted at a large high school, structured similarly to MCHS Smart, designed to introduce incoming students to the campus.

  • Group configuration: Students are grouped into units consisting of $8$ freshmen and $2$ senior student leaders, yielding $8 + 2 = 10$ individuals per group.

  • Interaction: Groups engage in planned activities throughout the day to foster peer connections.

• Sampling Protocol:

  • A random sample of $3$ complete groups is selected.

  • Every individual in each of the $3$ selected groups—including both freshmen and seniors—is surveyed.

  • Sample size: $3 \times 10 = 30$ total individuals surveyed.

• Evaluation of Sample Classifications:

  • Option A: It is a systematic random sample because the students were systematically divided into groups of $8$ freshmen (Incorrect; systematic sampling requires choosing every $k\text{th}$ subject from an ordered list).

  • Option B: It is a stratified random sample because all students were selected from some of the groups (Incorrect; stratified sampling requires selecting members from all strata).

  • Option C: It is a stratified random sample because some of the students were selected from each group (Incorrect; in this design, entire groups are chosen, not a subset from each group).

  • Option D: It is a cluster sample because all of the students were surveyed from the randomly selected groups (Correct; groups act as clusters where entire clusters are randomly sampled and every member within the selected clusters is evaluated).

  • Option E: It is a cluster sample because some of the students were selected from each group (Incorrect; cluster sampling surveys all units within selected clusters, not some from each group).

• Methodological Distinction Between Stratified and Cluster Sampling:

  • Stratified sampling: The population is partitioned into homogeneous groups (strata), and an $\text{SRS}$ of individuals is selected from every group (some from all).

  • Cluster sampling: The population is partitioned into heterogeneous, representative groups (clusters), a random sample of whole groups is selected, and all individuals within the chosen groups are evaluated (all from some).