Research: Independent Variables and Experimental Design

• PS219 is Research Methods in Psychology.
• Lecture 2 is about the basics of research and scheduled for January 14th, 2025.
• Rory Coyne is the PS219 Module Coordinator; [email protected].

Irish Cancer Society Summer Studentships 2025

• Applications are being accepted for the Summer Studentships Awards 2025.
• These studentships aim to offer undergraduate students an opportunity to conduct cancer research and work with researchers in high-quality environments.
• The program provides students with essential research experience early in their careers.
• Four summer studentship awards are available in total.
• Students can apply for either Translational Research or Survivorship Research.
• Eligibility for awards is determined by the proposed field of research.
• Virtual/desk-based projects are eligible if they meet all other criteria.
• The Irish Cancer Society Translational Research Summer Studentship 2025 allows undergraduates to undertake a translational cancer research project.
• Open to students enrolled in a degree, interested in translational, biomedical, or clinical-based cancer research.
• The application deadline is 3:00 pm, Thursday, February 13th, 2025.
• Project title example: "Influences on the uptake of and engagement with digital cancer rehabilitation services"

Topics Covered

• The scientific method of research.
• Independent and dependent variables.
• Different types of research.
• Scales of measurement.
• Measures of dispersion.
• Measures of central tendency.

The Scientific Method

• Identify a research question which may begin with previous research or theory, or a researcher's experience.
• Propose a hypothesis; a possible answer to the main research question.
• Choose a Research Method/Design, including questionnaires and interviews.
• Research designs may involve one point in time or many.
• The sample should be representative of the population of interest.
• Data are analyzed and published, which leads to new hypotheses and development of theories.

Hypothesis Testing

• The first step in psychological research is to define the variables being researched.
• The independent variable (IV) is identified as a possible cause of the phenomenon being researched.
• The IV is the variable being manipulated.
• Woodworth, 1938.
• The dependent variable (DV) is expected to change in response to changes in the IV.
• The DV is dependent on changes in the IV.
• A true experimental design means the IV is totally independent of any other variable.

Amount of water vs Amount of fertilizer example:

• Independent Variable: What is manipulated; Amount of water/fertilizer
• Little water and a lot of water are values of the Amount Of water independent value
• Little and a lot of fertilizer fertilizer are values of the Amount of fertilizer independent value
• Dependent Variable: Height of plant.
• Effect/Result shows that little water/fertilizer results in short plants while a lot of water/fertilizer results in a taller plant

Hypothesis Testing Continued

• Hypotheses: predictions about the outcome of an investigation/experiment.
• Two hypotheses constructed: the null hypothesis and the alternate (experimental) hypothesis.
• Null hypothesis (H0) always makes a statement of no (null) difference between the values of a population (e.g. means).
• H0 = There is no difference between the means for each of the two groups sampled.
• Ho = There is no difference in levels of anti-social behaviour between the couples that stay together V those that don't in Married At First Sight Australia.
• Alternate hypothesis (H1) provides the alternative to the null hypothesis.
• It always makes exactly the opposite statement about two values of a population (e.g. mean); there is a difference between them
• H1 = There is a difference between the means for each of the two groups sampled.
• H1 = There is a difference in levels of anti-social behaviour between the couples that stay together V those that don't in Married At First Sight Australia.

Hypothesis Testing Explained

• If the evidence is sufficiently strong, the alternate hypothesis may be supported and the null hypothesis rejected.
• In all other cases (i.e., when the evidence is not sufficiently strong), the alternate hypothesis is not supported.
• "Fail to reject" the null hypothesis.
• It is important to use this terminology when discussing psychology research.

Approaches to Research Design

• There are two main approaches to research design: Quantitative and Qualitative.
• Quantitative research methods gather numerical data or at least categories.
• Qualitative research methods gather non-numerical data.

Naturalistic Observation

• Naturalistic observation looks at a behaviour as it occurs in a natural setting.
• It makes no attempt at intervention on the part of the researcher.
• It is used when experiments are either difficult to carry out, or not as informative.

Survey Approach

• Can be quantitative or qualitative, but is mostly quantitative.
• It is often used when a researcher wants to gather data from a large number of people in a relatively short amount of time.
• It is very flexible and can be used to collect lots of different kinds of information.

Descriptive Designs

• These include naturalistic observation, surveys, and case studies (used to record the behaviour of a person, group or situation) .
• The aim is to observe and describe - the 'what' rather than 'why'.

Cross-sectional vs. Longitudinal Studies

• Cross-sectional study: a number of different variables are measured simultaneously for a sample of individuals.
• The strategy is to examine the extent to which these variables are associated with one another.
• Longitudinal Study: variables are measured at two or more points in time.
• Distinctions between IVs and DVs can get confusing in non-experimental designs.
• If comparing young people and older adults in terms of their computer literacy, age group is the independent variable, and computer literacy is the dependent variable.
• Doing so reveals if computer literacy changes depending on which age group the participant belongs to.
• Age cannot be manipulated by the researcher since it’s a fixed characteristic.

Experimental Approach

• Experimental approach features tight control by the researcher.
• Less likely to have other variables.
• True Experimental Design includes random assignment and a control group.
• Double blind experiment means both participant and researcher are blind to the condition the participant is assigned to.
• Experimental approach includes dividing participants randomly into either experimental or control groups.
• The experimental group receives the experimental procedure.
• The control group receives the control procedure which is different in some way.
• The differences between the two groups are assessed on another variable.

True vs. Quasi Experiments

• A distinction can be drawn between 'true' and 'quasi' experiments.
• A true experiment can identify a causal relationship between an independent and dependent variable.
• To qualify as a true experiment: random assignment to conditions, equal treatment of subjects, and a control condition.
• Experimental Condition: The 'intervention' condition represents changes to IV values away from the control condition, to see if any changes occur in the dependent variable.
• Control Condition: represents the 'nil intervention'

Quasi Experiment

• A quasi experiment will not permit a causal relationship to be established.
• They enable the collection of data which could not be acquired by means of a true experiment.
• A quasi-experimental design is used where the independent variable cannot be manipulated by the researcher for ethical reasons such as self esteem.
• Data could be collected on self-esteem and people assigned to groups as a result.

Research Designs

• Cross-sectional design: simple and naturalistic, but poor at showing causality.
• Longitudinal design: complex to implement, but improves confidence in causality.
• Experimental design: risks are somewhat artificial and it is complex to implement, but provides the strongest evidence of causality.

Choosing a method depends on:

• The information desired and the aims of the study.
• The nature of the phenomenon.
• Is it feasible to collect data and would it be valid/reliable?
• Is it ethical to conduct the study?
• The cost of the design .
• How much current scientific theory and literature exist on the topic.

Scaling (N.O.I.R.)

Four types of data measurement:

• Nominal: example - favourite food/gender/type of pet. Data can only be classified. Ordinal: example - placement in a race. Can be ranked in some kind of way, but degrees of difference between ranks are ignored.
• Interval: example - temperature. There are meaningful differences between values, but has no true zero.
• Ratio: example - weight/age. This has a meaningful zero and ratio between values.

Ratio = Absolute zero Interval = Distance is meaningful Ordinal = Attributes can be ordered Nominal = Attributes are only named; weakest

Types of Measurement

• Nominal variables: the variable consists of named categories. The categories have no mathematical properties.
• Ordinal (rankable): The scores indicate only rank order in terms of size and it is incorrect to calculate means on the scores.
• Interval or equal interval: The steps between the scores are equal in size though there is no proper zero point. The scores can be added, means calculated, etc.
• Ratio is the same as 'Interval' measurement but the scale has a proper zero point. Ratios can be calculated as consequence.

Key Terms

• Raw Data: Consists of the measurements of a variable as collected directly from source during an investigation and before any statistical processing takes place.
• Statistics: Values computed from a sample (e.g. mean, median, mode, SD, etc.).

Measures of Central Tendency

• The Mean: The arithmetic average of a set of data, represented in statistical notation by the symbol X.
• The Median: The positional average of a set of scores which represents their central tendency in a very literal way (i.e. the midpoint when the scores are placed in ascending order).
• The mode: The value that occurs most often in a set of data

Distributions of Data

• A distribution: name given to any collection of data once the values have been organised in a way which enables the 'shape' of the data to be seen.
• Cumulative frequency: allows one to ascertain where a particular value lies in relation to the sample as a whole.
• A frequency distribution: obtained by counting the number of times each of the possible values occurs in the set of score (represented on a graph or chart).
• Negative Skew: More values on the right-hand side.
• Normal distribution: Perfect bell-shaped curve.
• Positive Skew: More values on the left-hand side.
• Real-life data: often does not follow a normal distribution.

Examples

• Yearly income (thousand dollars) is usually an example of Positive Skew.
• Deaths in Australia in 2012 is an example of Negative skew.
• Mean: ~77 years , Mode: 87 years , Median: "81 years

Measures of Spread

• Range: The difference between the largest and smallest score of a variable.
• Variance: Measures how far the data is spread out from the mean. If the variance is small, the data is concentrated around the mean.
• Standard deviation: Conceptually, the average amount by which scores differ from the mean.
• Standard deviation root of variance.
• Standard deviation - same units as the original values whereas the variance is expressed in squared units.

Quiz Answers

• If I add more milk, the tea will get a lighter color.

• The two variables: 1. Amount of milk 2. Color of the tea

• The independent variable = Amount of milk

• Mosquito repellent is sprayed on one arm and the other arm is not sprayed. The number of mosquito bites is counted after 2 hours.

• The two variables: 1. Mosquito bites. 2. Repellant sprayed.

• The independent variable = Repellant sprayed.

• Javier sleeps 9 hours the night before his science exam. His twin sister has the same exam and sleeps 6 hours the night before the test. Afterwards, they compare their scores.

• The two variables: 1. Hours slept. 2. Exam scores.

• The dependent variable = Exam scores.

• New houses are spayed on the outside with an insulating paint. After a year, researchers determine how much energy each house used.

• The dependent variable = how much energy each house used.

Recommended Reading

• Introduction to Research Methods in Psychology Third edition By Dennis Howitt and Duncan Cramer
  • Chapter 1 - 'The Role of Research in Psychology'
  • Chapter 2 - 'Aims and Hypotheses in Research'
  • Chapter 3 - 'Variables, Concepts and Measures'
• Introduction to Statistics in Psychology 5th edition By Dennis Howitt and Duncan Cramer
  • Chapter 2 - 'Some basics'
  • Chapter 4- 'Describing Variables Numerically'