VIDEO 3 Jenis uji statistik AWAL

Summary

This video lecture explains the importance of choosing the right research design and statistical tests for research. It covers various types of research, including quantitative and experimental designs, and introduces the concepts of univariate, bivariate, and multivariate data analysis. The speaker emphasizes the distinction between numerical and categorical data, highlighting how these differences influence the choice of statistical tests. The lecture also provides a step-by-step guide on how to conduct bivariate analysis, focusing on understanding relationships between independent and dependent variables, essential for students preparing theses or dissertations.

Takeaways

• 😀 Understanding the type of research is essential for determining the appropriate statistical test to use.
• 😀 Research methodology requires clarity on three key elements: the type of research, the design, and the approach.
• 😀 A research design example for quantitative research could be 'non-experimental with a cross-sectional approach'.
• 😀 For experimental research, a design like 'quasi-experimental with a non-equivalent control group' is commonly used.
• 😀 Research design directly influences the choice of statistical analysis techniques that will be applied.
• 😀 Data analysis can be univariate (one variable), bivariate (two variables), or multivariate (multiple variables).
• 😀 Most undergraduate theses typically only require univariate or bivariate data analysis, while multivariate analysis is used in higher-level research (e.g., master’s or Ph.D. theses).
• 😀 Univariable analysis involves a single variable, while bivariate analysis compares two variables: an independent variable and a dependent variable.
• 😀 Multivariate analysis explores the relationships between multiple variables, whether independent or dependent.
• 😀 Data types can be categorized into numerical (measured data) and categorical (qualitative data), each requiring different analysis methods.

Q & A

What is the importance of determining the research design in a study?

Determining the research design is essential because it influences the choice of statistical tests. A proper design ensures that the data analysis will be valid and appropriate, helping researchers avoid errors in test selection.

What are the three main components that should be included in the research design section?

The three main components that should be included in the research design section are: 1) The type of research (e.g., quantitative or qualitative), 2) The research design (e.g., experimental, observational), and 3) The approach or method used (e.g., cross-sectional study, experimental design).

How does the research design influence statistical analysis?

The research design determines the appropriate statistical analysis by defining the type of data and variables involved, which in turn dictates whether bivariate, multivariate, or univariate analyses are suitable.

What is the difference between univariate, bivariate, and multivariate analysis?

Univariate analysis involves the examination of a single variable, bivariate analysis involves two variables (typically one independent and one dependent), and multivariate analysis examines multiple variables, either independent or dependent, simultaneously.

What type of data is considered 'numerical' and how does it differ from 'categorical' data?

Numerical data consists of values that are derived from measurements (e.g., weight, temperature) and are represented by numbers. Categorical data, on the other hand, consists of labels or categories, such as 'yes/no', 'anemia/non-anemia', or educational levels like 'elementary', 'middle school'.

Why is it important to understand the difference between numerical and categorical data?

Understanding the difference is crucial because it affects the choice of statistical analysis. Numerical data requires techniques like mean, median, or regression analysis, while categorical data is analyzed using methods such as chi-square tests or frequency counts.

What does 'bivariate analysis' involve in research?

Bivariate analysis involves analyzing the relationship between two variables, typically one independent and one dependent variable. It helps determine if and how the two variables are related.

Can you provide an example of categorical data in a study?

An example of categorical data would be the classification of patients as 'anemic' or 'not anemic' based on their hemoglobin levels. Though the hemoglobin levels themselves are numerical, they are categorized into groups for analysis.

What is the difference between experimental and non-experimental research designs?

Experimental research designs involve manipulating an independent variable to observe its effect on a dependent variable (e.g., controlled trials), while non-experimental designs observe variables without manipulation, such as observational or survey-based studies.

What is the role of the 'methodology section' in a research paper?

The methodology section explains how the research was conducted, including the research design, data collection methods, and data analysis procedures. This section provides transparency and allows others to replicate the study if necessary.