Quantitative Methods M.2 - The Scientific Method

What are the 5 phases of the empirical cycle (the scientific method)?

1. Observation: observing a relation between some things, which sparks the idea for a hypothesis

2. Induction: takes a hypothesis and infers that the statement is true every time, turning that hypothesis into a general rule

3. Deduction: infers that the relation specified in the general rule should also maintain itself in new, specific instances

4. Testing: involves collecting new data and comparing it to the original prediction w/ statistics

5. Evaluation: results are interpreted in terms of what the hypothesis says, the hypothesis is either provisionally supported, adjusted, or rejected (the latter is actually rare)

What are the main criteria with which to decide if a study is convincing?

1. Reliability: the study is repeatable and leads to similar results

2. Validity: the relation between properties accurately reflects reality

What are the main types of validity?

• Construct validity: happens if the properties / constructs are measured and manipulated accurately. It is a prerequisite for the two other types of validity.

  • Internal validity: the observed effect is actually due to the hypothesized cause (causation > correlation)

  • External validity: the hypothesized relationship also holds in other settings, groups and environments

What are the main prerequisites for a causal relationship between properties?

1. The cause and effect are connected; we can trace the effect back to the cause

2. The cause comes before the effect

3. The cause and effect occur together consistently

4. Alternative explanations can be ruled out

What are the main threats to internal validity?

1. Maturation: a natural change occurring in the property itself

   1. To prevent this, we can introduce a control group that is measured at the same time but not exposed to the hypothesized cause

2. Selection: the difference in subject characteristics between groups that may be affecting the final result

   1. Randomization, or the process of assigning participants to different groups in a study by chance, can solve this problem

3. Selection by maturation: the combined threat of selection and maturation, which occurs when groups systematically differ in their rate of maturation

What are the main threats to internal validity that are associated with the tools used to measure/manipulate the constructs?

1. Low construct validity: the tools contain a systemic bias, or measure another construct or property entirely

2. Instrumentation: an instrument used to measure a property/construct is changed during the course of the study

   1. Both low construct validity and instrumentation can be eliminated by using valid instruments and valid manipulation methods, and by using them consistently

3. Testing: taking a test can change participants’ behavior, which may affect the study’s results

   1. Can be eliminated using a special design that includes groups which are exposed to a pre-test and groups which are not