While descriptive statistics are used by a researcher to organise, summarise and describe data, inferential statistics are used by a researcher to interpret the data and make inferences from observations of a sample in an experiment to generalise to a wider population.

Stratified Sampling: Represents the general population on a certain characteristic. This prevents bias in the sample and enables the researcher to generalise the findings to the wider population.
In some experiments, the sample needs to be chosen to represent the general population on a certain characteristic, e.g. gender or age. Therefore, the researcher makes sure that the proportion of people in the sample with the particular characteristic is the same as the proportion of people in the general population with that characteristic.
When subjects within each stratum are selected in a random way, it is called a stratified random sample. This method prevents biases in the sample and enables the researcher to generalise the findings to the wider population.

Random Sampling: This involves the selection of subjects in such a way, that each has an equal opportunity to be chosen as a participant in the experiment. A number of methods can be chosen such as: picking names out of a hat or choosing every third person on an alphabetical list.
Ensures that subjects have an equal opportunity to be chosen as participants in the experiment. Usually, yields a fairly representative result.
If the sample is large enough, random sampling usually yields a fairly representative result.

Dependent Variable (DV): This is the variable which is being observed in the experimental situation to see the effect of the independent variable. Think of it as the outcome or results of both the experimental and control groups.

Control Group: The group of subjects which is matched in all ways possible with the subject characteristics and conditions of the experimental group except that it is not exposed to the independent variable. This group serves as a baseline measure for comparison with the experimental group. The control group helps the researcher to determine if the independent variable did, in fact, cause an effect on the dependent variable.

Independent Variable (IV): This is the variable that is manipulated or varied by the experimenter. It is applied to the experimental group and is known as the cause of the results.

Experimental Group: The group of subjects which is matched in all ways possible with the subject characteristics and conditions of the control group but is also exposed to the independent variable.
The purpose of the experimental group is to provide a measure of the influence in the dependent variable produced by the intervention.

Placebo Effect: A placebo effect is when there are changes in the subjects' behaviour caused by the belief that they received an active drug or treatment when, in fact, the drug was a neutral substance. This inert or neutral 'treatment' is called a placebo and is used to equalise the expectations of the control and experimental groups.
For example, if you wanted to test the effects of a new drug which was supposed to improve memory, you could have two groups. The experimental group would be given the drug in tablet form and the control group would be given tablets that looked the same as those received by the experimental group, except that they contained glucose (a placebo). Both groups would then be tested on a memory task. In this experiment, the subjects feel that they have been treated equally and any difference in the groups would most likely be due to the drug.
In another experiment, three groups could be tested where group one was given the drug, group two, a placebo, and group three nothing. If the results of groups one and two were very similar and much higher than group three, then this is an example of a placebo effect.

Double-blind: In a double-blind experiment, the data collection is carried out by a third party who is told what to do but has no knowledge of the expected outcomes or which group is which.
Thus neither the subjects nor the experimenter can bias the experiment.

Single-blind: In a single-blind experiment, the subjects do not know whether they are in the control group or experimental group condition.
This prevents the subjects' knowledge or expectations of the experiment impacting on the results.

Matched-subjects Design: As well as utilising a repeated-measures design to minimise the effects of extraneous variables, researchers often choose to conduct a matched-subjects design. A matched-subjects design also attempts to obtain equivalence between groups thus increasing the sensitivity of the experiment. It is where subjects are matched or equated on one or more variables (other than the independent variable) assumed to have an effect on the dependent variable. Then, whenever possible, the subjects are randomly assigned to conditions. In this way, the sensitivity of the experiment is increased.
For example, a researcher may identify the variable of subject intelligence as having an impact on the dependent variable. All the subjects would be matched on intelligence level, holding this variable constant and therefore controlling for it in all groups. The subjects would then be randomly assigned to experimental conditions.
A matched-subjects design can also mean that each subject has a matched subject in each of the other conditions so that the groups are correlated. Therefore, rather than using the same subjects in each condition as in the repeated-measures design, the matched-subjects design uses different subjects in each condition who have been matched as much as possible on identified variables.

Independent-groups Design: This research design is also known as a between-subject design as it has different subjects in each group. The separate or independent groups of subjects receive different levels of the independent variable, so there is no chance of one treatment contaminating the other. In its simplest form, the experimental group receives the independent variable and the control group does not.
However, the independent-groups design has the potential of introducing variables other than the independent variable because of differences among the subjects in the groups. These extraneous variables may confuse the clear relationship between the independent variable and the dependent variable. Therefore, to minimise the confounding of subject characteristics with the independent variable, subjects need to be either randomly assigned to the different conditions or matched between groups on personal characteristics. These procedures ensure the formation of equivalent groups of subjects.

Repeated-measures Design:In order to control for variations in subject characteristics (which may influence the dependent variable), a within-subjects design may be used called a repeated-measures design. This is where the subjects within an experiment are tested or measured more than once, i.e. they are repeatedly measured. The same subjects are used in each condition, receiving all levels of the independent variable thereby guaranteeing equivalence of groups at the beginning of the study.
For example, subjects may be given a pre-test and then asked to complete a post-test. In this way, the subject characteristics remain constant.
Another example is if a researcher wanted to study attention span for complex and simple visual stimuli. The researcher may test the same children who would be exposed to both the complex and the simple visual stimuli conditions. The attention span of all children would be tested under each of the two conditions, hence the subjects are repeatedly measured. This ensures that the two groups are equivalent in subject characteristics, minimising the influence of extraneous variables.

Definition of Operational Hypothesis: The IV and the DV in an experiment need to be operationalised. This refers to the precise way in which the variable has been measured. It is a statement of how the variable can be expressed as a quantity.
Before collecting any data, a hypothesis needs to be formulated. A hypothesis is an 'educated guess' about the relationship(s) between two or more variables which can be scientifically tested. Psychologists use the word 'educated' as the hypothesis is derived from some theory or previous research which educates the researcher when formulating a hypothesis for his/her study. A hypothesis is also a 'guess' because it is not a statement of fact or certainty and could be supported or refuted through the process of data collection.
Initially, the hypothesis may be stated in more general terms, e.g. 'that stress increases blood pressure'. This form is too broad to be tested statistically and must be expressed more precisely.
When designing the research, the hypothesis needs to be operationalised, i.e. the variables must be clearly defined and measurable. The operational hypothesis should include both the independent and dependent variables and predict a direction of the changes in the dependent variable.