Quantitative research operates with variables. There are three types of quantitative research;

• Experimental studies: The experiment in its simplest form includes one independent variable (IV) and one dependent variable (DV), while the other potentially important variables are controlled. The experimental study is the only method that allows cause-and-effect inferences.

• Correlational studies: There are no IVs or DVs. Variables are measured and the relationship between them is calculated.

• Descriptive studies: Relationships between variables are not investigated, and the variables are approached separately.

Experimental Studies

Sampling in the experiment

Several sampling techniques can be used in an experiment. The choice depends on the aim of the research.

• Random sampling: Every member of the target population has an equal chance of becoming a part of the sample. The results achieved by this sampling method are easily generalizable.

• Stratified sampling: First, essential characteristics for the study are decided, then the distribution of these characteristics in the general population is studied. Finally, participants are recruited in a way that keeps the proportions observed.

• Convenience sampling: Easily available participants are recruited. It is quicker and financially easier than other methods. This method is useful when wide generalization is not the primary goal.

• Self-selected sampling: It takes place by recruiting volunteers. It is quick, easy, and has a wide coverage. The limitation of this method is its representativeness.

Experimental Designs

The organization of groups and conditions in an experiment is known as the experimental design, and there are three basic types of it.

• Independent measures design: Involves random allocation of participants into groups and a comparison between these groups. The IV is manipulated and there can be more than one IV.

• Matched pairs design: Instead of completely random allocation, matching is used to form the groups. It is used to make sure that the groups are equivalent in a specific variable.

• Repeated measures design: The goal is to compare conditions rather than groups of participants. The same group of participants is exposed to two (or more) conditions, and the outcomes are compared. The problem with repeated measures design is order effects, which change depending on which condition comes first. They may appear due to practice and fatigue. To counteract this, another group of participants is used and the condition order is reversed.

Types of Validity

• Construct validity: Is high when the making of unmeasurable variables measurable is justified and provides sufficient coverage. Characterizes the quality of operationalizations.

• Internal validity: Is high when CV has been controlled and we’re sure that it was the change in the IV that caused the DV to change. Characterizes the quality of the experiment.

• External validity: Characterizes generalizability of findings in the experiment. There are two types of it;

  1- Population Validity: generalizability from the sample to the target population, high when sample is representative

  2- Ecological validity: generalizability of the experiment to other settings or situations

Bias in Experimental Research

Confounding factors that may influence the cause-and-effect relationship between the IV and the DV

1. Selection: Mistakes in sampling and creating groups

2. History: The outside events that happen to a participant/ group (ex. noise coming from outside in a memory test, one group is closer to the noise)

3. Maturation: Participants going through natural development (ex. child participants may grow in between two experiments)

4. Testing effect: Doing a test for the second time affecting the results, thus the researcher not knowing if the results changed because of the training given or familiarity

5. Instrumentation: The instrument measuring the DV changing slightly (ex. the observer being more tired during one of the experiments)

6. Regression to the mean: The DV being extremely high or low (when a person retakes a test, their score tends to get closer to the average score)

7. Experimental mortality: Some participants dropping out during an experiment, may become a problem if dropouts aren’t random (ex. if the dropouts are higher in the experimental group, the experiment cannot go on)

8. Demand characteristics: Participants understanding the aim of the study and altering their behavior accordingly (in order to eliminate this, the experiment should be a blind experiment)

9. Experimenter bias: The researcher unintentionally affecting the results of a study (in order to eliminate this bias, the experiment should be a double blind experiment)

Quasi-experiments vs. true experiments

In quasi-experiments, there is no random allocation or manipulation of the IV unlike true experiments. In both experiment types, the third variables are controlled. Pre-existing differences which cannot be manipulated are the main reasons for quasi-experiments. (age, gender, age, anxiety, occupation…)

Field experiments and natural experiments

Field experiments are conducted in a real life setting and the IV’s are manipulated but since participants are in their natural setting CV’s cannot be controlled. Natural experiments are also conducted in a natural environment but here the researcher has no control over the IV’s and CV’s.

Correlational studies

No variable is manipulated by the researcher so causation cannot be inferred. Two or more variables are measured and the relationship between them is mathematically quantified. This relationship can be; no relationship, negative correlation, and positive correlation.

Limitations/ biases of correlational studies

1. The third variable problem: The possibility that there is a third variable that correlates with both X and Y and explains the correlation between them.

2. Curvilinear relationships: Sometimes variables are linked non-linearly.

3. Spurious correlations: When there are multiple correlations between multiple variables, there is a possibility that some of the correlations are from random chance.