BOLD Educational Software Scientific Method

The scientific method is the process by which researchers attempt to construct an accurate (reliable and consistent) representation of the world.

The scientific method consists of a number of steps that correspond to the chapters in your dissertation or research proposal or the parts of a research article in a journal:

1. Explore a subject area and define a problem. This is covered in the Chapter 2 literature review as well as the background of the problem in Chapter 1.

2. Define a research question: This is Chapter 1 of your dissertation.
It is supported by Chapter 2 (the literature review), which provides the background of the study.

3. Formulate Hypotheses (Restate your research question as a statement, and make a prediction of the outcome). This is also part of Chapter 1.

4. Design the Study: What procedures will you use to collect and analyze data? This is laid out carefully in Chapter 3 of your dissertation.

5. Collect and Analyze data: This is Chapter 4 of your dissertation. Note: the analysis conducted in Chapter 4 MUST match the analysis laid out in Chapter 3, so plan the analysis carefully!
Pilot studies are often a good idea so you can discover design flaws before you conduct the actual analysis.

6. Draw conclusions: What have you learned from the study? This is Chapter 5 of your dissertation.

When conducting a research study, you will need to know three things:

1. What is the research question you are trying to answer?
This must be a clear question that can be answered by a carefully designed study.
The research question comes DIRECTLY from the research purpose statement,which is in turn derived directly from the problem statement.

2. What are the data?
Every research question will contain one or more variables - the data are the pieces of information you collect on those variables and must be measured in a precise way.

3. What is the design of the study?
The design consists of two things:

1. Methodology (experimental, quasi-experimental, nonexperimental), and
2. The statistical procedure (e.g., correlation, regression, ANOVA).

Data
Constants: characteristics or values that do not change; same value for every person or object.
If you only include female participants in your study, you are holding the variable GENDER constant.

Variables: characteristics or values of people, places, events, or things that do change from person, place, event, or thing within the sample.
If both male participants and female participants are included in your study, then GENDER is a variable.

There are, broadly speaking, two types of quantitative statistics:
1.    Descriptive Statistics describe a single set of numbers or data; that is, descriptive statistics summarize the numbers and depict patterns formed by those numbers.
2.    Inferential Statistics examine the relationship between two or more variables. Inferential statistics are used to analyze relationships between variables and answer the researcher's questions.
Inferential statistics are used to test hypotheses.

Types of Quantitative Research Questions

Descriptive Statistics

1.       Descriptive (e.g., What is the mean grade point average of students in this school?)
Describe/Summarize a Set of Data

Sample Descriptive Quantitative Research Questions:

1.       What percentage of participants falls in each category?
(Present frequency tables)
2.       What percentage of participants received each score for [name variable]?
(Present frequency tables)
3.       What is the distribution of [name variable]?
(Present histograms, bar charts, or other graphic illustrations)

Inferential Statistics

Draw inferences

1.      Correlational (Relational) - Nonexperimental (e.g., Are the number of churches in midwestern towns related to the number of bars in each town?)

2.     Experimental/Quasi-experimental (e.g., Are students who take distance education classes learning as much as students who take the same class on campus?)

Sample Inferential Quantitative Research Questions:

1.       Is there a significant relationship between [variable 1] and [variable 2]?
(Pearson Product Moment Correlation Coefficient or Spearman Rank Correlation)
2.       Is there a significant difference between [variable 1] and [variable 2]?
(t-test or analysis of variance)