ANOVA

Definition:
- ANOVA is used to test differences between the means of three or more groups simultaneously.
- It assesses whether group means differ on a particular score or variable.
Test Statistic:
- The statistic used for ANOVA is the F-test.
- Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Simple Analysis of Variance (One-Way ANOVA):
- Applied when one factor or treatment variable is being investigated across more than two groups (levels).
- Focuses on the influence of one independent variable on a dependent variable.
Factorial Design:
- Used when exploring multiple treatment factors.
- It allows for the analysis of two or more independent variables.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Definition:
- Factorial Design ANOVA examines the effects of more than one treatment or factor simultaneously.
- It involves multiple independent variables and one dependent variable.
Example:
- A 3x2 factorial design, which means three levels of one factor and two levels of another.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Scenario: Evaluation of preschool hours on gender:
- Parameters:
- Gender: Male, Female
- Hours in Preschool: 5 hours/week, 10 hours/week, 20 hours/week
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Understanding Variation:
- A significant variation between samples suggests differences between groups, while little variation within samples indicates consistency among them.
- Assumption: Variation relates to the group factor (e.g., treatment, gender).
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Null Hypothesis (H0):
- Specifies that the means of the groups are equal:
- H0: 1 = 2 = 3
Research Hypothesis:
- Indicates that at least one group mean is different from the others.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Source Table Example:
- The breakdown of sum of squares (SS), degrees of freedom (df), mean square (MS), and F-value:
- Between Groups:
  - SS: 1,133.07
  - df: 2
  - MS: 566.54
  - F: 8.799
- Within Groups:
  - SS: 1,738.40
  - df: 29
  - MS: 64.39
Note: The F value for two groups is equivalent to the square of the t-value (F = t²).
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Calculating Degrees of Freedom:
- Numerator: Number of groups minus one: k - 1
- Example with 3 groups: 3 - 1 = 2
- Denominator: Total number of observations minus the number of groups: N - k
- Example with 30 participants: 30 - 3 = 27
- Representation: F(2, 27)
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Example Statement:
- For F(2, 27) = 8.80, p < .05
- Breakdown of the statement:
  - F indicates the test performed.
  - Degrees of freedom: 2 (between-group) and 27 (within-group estimates)
  - Obtained F-value: 8.80
  - p-value significance: less than 0.05, indicating statistical significance.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Data Requirement:
- Description on how to prepare data for a single-factor ANOVA.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Importance:
- If a significant difference is found, post hoc analysis is necessary to determine which specific groups have different means.
- Conducting multiple t-tests is not appropriate after finding a significant ANOVA result.
Post Hoc Procedure:
- Tukey’s Honest Significant Difference (HSD) is typically utilized for follow-up comparisons among group means.
Note: Detailed methodologies for this are covered in Chapter 18.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Functionality:
- Used when testing more than one factor or independent variable.
- This methodology provides insights into not only the individual effects of each factor but also their interactions.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Two Types:
- Without Replication:
- One factor is not tested multiple times across the same subjects.
- With Replication:
- Multiple measurements are taken on the same subjects across one factor.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Example Questions Addressed:
- Is there a difference in weight loss effects between high-impact and low-impact exercise programs?
- Is there a difference in weight loss effects based on gender (male vs. female)?
- Do the effects of the exercise program differ between genders?
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Definition:
- A main effect occurs when an independent variable significantly impacts the outcome variable.
Source Table:
- ANOVA summary table that presents sources of variance.
Interaction Effect:
- This refers to the effect that occurs when the impact of one independent variable varies depending on the level of another independent variable.
Source: Salkind, Statistics for People Who (Think They) Hate Statistics, Excel 2016, 4th Edition, SAGE Inc. © 2016

Null and Research Hypotheses:
- These are akin to the hypotheses explored in standard ANOVA, focusing on the equality of means and the differences among them.

Source Table Example:
- Illustration of Main Effects:
- Impact:
  - Sum of Squares: 429.025
  - df: 1
  - Mean Square: 429.025
  - F: 3.011
  - Significance: .091
- Gender:
  - Sum of Squares: 3222.025
  - df: 1
  - Mean Square: 3222.025
  - F: 22.612
  - Significance: .000
- Interaction (Impact x Gender):
  - Sum of Squares: 27.225
  - df: 1
  - Mean Square: 27.225
  - F: .191
  - Significance: .665
- Error:
  - Sum of Squares: 5129.700
  - df: 36
  - Mean Square: 142.492

Purpose:
- Visual representation of main effects (without interaction) to aid in understanding.

Source Table Example:
- Breakdown for interaction effects:
- Treatment:
  - Sum of Squares: 265.225
  - df: 1
  - Mean Square: 265.225
  - F: 2.44
  - Significance: .127
- Gender:
  - Sum of Squares: 207.025
  - df: 1
  - Mean Square: 207.025
  - F: 1.908
  - Significance: .176
- Interaction (Treatment x Gender):
  - Sum of Squares: 1050.625
  - df: 1
  - Mean Square: 1050.625
  - F: 9.683
  - Significance: .004
- Error:
  - Sum of Squares: 3906.100
  - df: 36
  - Mean Square: 108.503
- Total: 224321.000

Purpose:
- Visual representation of interaction effects to interpret the data effectively.

Dialog Box:
- Guide on how to set up and analyze two-factor ANOVA with replication using Excel.