Central Tendency

Chapter 3: Descriptive Statistics and Central Tendency

Introduction to Central Tendency

  • The average score represents a distribution and summarizes data.

  • Central tendency is the statistical measure that identifies a single value as representative of the entire distribution.

Purpose of Central Tendency

  • Describes Population or Sample: Simplifies data for understanding.

  • Facilitates Comparison: Allows comparing different groups.

Methods for Measuring Central Tendency

  1. Mean

  2. Median

  3. Mode

The Mean

  • Definition: The arithmetic average of a set of values.

  • Symbol for Population Mean: μ

  • Symbol for Sample Mean: M or X

Calculating the Mean

  • Example: For scores 3, 7, 4, 6 (N = 4), the mean can be calculated.

  • For a family of 5 members, total age = 100; Mean = Total Age / Number of Members.

  • For class UTS scores with total scores listed, find Mean similarly.

The Mean as a Balance Point

  • Determines a balance point of scores in a sample.

  • Example: For scores (1, 2, 6, 6, 10) identify where the mean lies.

The Weighted Mean

  • Used when combining multiple samples: e.g., Sample 1 (n = 12, M = 6) and Sample 2 (n = 8, M = 7).

  • Calculate the mean of the combined group.

Computing the Mean from Frequency Distribution

  • Using the formula: M = Σ f.Xc / N

  • Example frequency table provided to illustrate the concept.

  • Precise calculation methods are needed when data is grouped or extensive.

Characteristics of the Mean

Changing a Score

  • What happens: If you change one of the scores in a dataset, the mean will also change, but the amount of change depends on how much the score was changed.

  • Why: The mean is based on the total sum of all scores divided by the number of scores. So, altering a single score changes the total sum, which directly affects the mean.

2. Introducing or Removing a Score

  • Introducing a Score: Adding a new score to the dataset affects the mean depending on the value of that score:

    • If the new score is higher than the current mean, the mean will increase.

    • If the new score is lower than the current mean, the mean will decrease.

  • Removing a Score: Removing a score works the same way but in reverse:

    • If the removed score is higher than the mean, the mean will decrease.

    • If the removed score is lower than the mean, the mean will increase.

3. Adding or Subtracting a Constant

  • What happens: If you add or subtract the same constant to every score in the dataset, the mean will also increase or decrease by that same constant.

  • Why: The total sum of the scores shifts by that constant multiplied by the number of scores, and since each score changes equally, the mean shifts by the constant.

Example:

  • Original scores: 5,6,75, 6, 75,6,7, mean = 666.

  • Add 222 to each score: 7,8,97, 8, 97,8,9, new mean = 888 (mean increased by 222).

4. Multiplying or Dividing by a Constant

  • What happens: If you multiply or divide all the scores by the same constant, the mean will also be multiplied or divided by that constant.

  • Why: The total sum of scores changes proportionally, and since the number of scores stays the same, the mean scales up or down by the same factor.

Example:

  • Original scores: 5,10,155, 10, 155,10,15, mean = 101010.

  • Multiply by 222: 10,20,3010, 20, 3010,20,30, new mean = 202020 (mean doubled).

Benefits of Using the Mean for Frequency-Grouped or Continuous Data

  1. Utilizes All Data Points:

    • Considers every value in the dataset, making it a comprehensive summary.

  2. Represents the Total Distribution:

    • Reflects the overall pattern and central tendency of the data.

  3. Compatible with Further Statistical Analysis:

    • Used in calculations for standard deviation, variance, and other advanced techniques.

  4. Efficient for Grouped Data:

    • Allows calculation using midpoints and frequencies, simplifying analysis for large datasets.

  5. Balances Variability:

    • Smooths out differences by taking both small and large values into account.

  6. Ideal for Symmetrical Distributions:

    • Accurately represents the central value in datasets with balanced distributions.

  7. Mathematically Reliable:

    • Provides consistent results, especially in repeated sampling or comparisons.

  8. Widely Understood:

    • Familiar to most users, making interpretation and communication of results easier.

→  the mean uses every score in the distribution, it typically produces a good representative value. The mean has the added advantage of being closely related to variance and standard deviation, the most common measures of variability.

The Median

  • Definition: The point where 50% of scores fall below.

  • Alternative term: P50; Symbol: Mdn.

Calculating the Median

  • Method 1 (Odd N): Order scores and find the middle score.

  • Method 2 (Even N): Average the two middle scores.

Continuous Variables

  • Consideration of score intervals for precise median calculation.

  • Example using identical middle scores (1, 2, 2, 3, 4, 4, 4, 4, 4, 5).

Interpretation of the Median

  • Addressing issues of interpreting distribution scores as exact values.

When to use Median

  • The median is particularly useful in situations where the data set contains outliers or skewed distributions

  • In these cases, the median helps to avoid misleading conclusions that can arise from extreme values, allowing for more accurate insights into the typical score of the data set.

  • Undetermined values

  • Open-ended distribution

  • Ordinal Scale

The Mode

  • Definition: The score that occurs most frequently in a dataset.

  • In grouped data, taken as the midpoint of the most common class interval.

  • When to use Mode → Nominal scales, discrete variables, describe shapes

Choosing a Measure of Central Tendency

  • Evaluation of whether to use mean, median, or mode is dependent on specific factors:

  1. Nature of the data.

  2. Distribution shape.

Reporting Measures in Literature

  • Example: Reporting means for treatment and control groups with respective M and Mdn values.

Central Tendency and Distribution Shape

  • Relationship between mean, median, and mode depends on distribution shape.

Symmetrical Distribution

  • Right and left sides are mirror images.

  • Mean and median are identical.

Rectangular Distribution

  • Has no mode

  • Has Mean & Median, located in the center of the value

Bimodal Distributions

  • Mean and median-centered, modes on either side.

  • Implications for interpretation of data.

Positively/Negatively Skewed Distributions

  • Analysis of how skew affects measures of central tendency.