Psychological skills

10.1 · Free notes

Planning and designing investigations: aims, hypotheses, variables, designs, controls and pilot studies

Key points

• An aim is a general statement of what the researcher intends to investigate and why. A hypothesis is a precise, testable prediction written in operationalised terms.
• A directional (one-tailed) hypothesis predicts the direction of the difference or correlation, and is used when earlier research or theory supports it. A non-directional (two-tailed) hypothesis predicts a difference or relationship without stating the direction.
• The null hypothesis states that there is no significant difference or relationship, and that any observed pattern is due to chance. It is retained or rejected after a statistical test.
• The independent variable (IV) is manipulated by the researcher. The dependent variable (DV) is measured. Operationalise both, for example 'recall = number of words correct out of 20 in 2 minutes'.
• Extraneous variables (e.g. noise, participant mood) may affect the DV but do not vary systematically with the IV. Confounding variables vary systematically with the IV, so the cause of any change in the DV cannot be identified.
• Independent groups uses different participants in each condition. It has no order effects, but participant variables can differ between groups, which is controlled by random allocation.
• Repeated measures uses the same participants in every condition. It controls participant variables but creates order effects (practice, fatigue) and demand characteristics. Matched pairs pairs participants on a relevant variable, and then one of each pair goes to each condition. It is time-consuming and matching is never perfect.
• Counterbalancing (ABBA) means half the participants do A then B and half do B then A, which spreads order effects evenly. With k conditions, complete counterbalancing needs k! orders (3 conditions = 6 orders).
• Key controls are standardised procedures and instructions, controlling situational variables (temperature, noise, time of day), and random allocation. A single-blind design hides the condition from participants, and a double-blind design hides it from the researcher too, reducing demand characteristics and investigator effects.
• Laboratory experiments have high control but low ecological validity. Field experiments manipulate the IV in a natural setting, so they have higher ecological validity but less control. Natural/quasi experiments use an IV that is not manipulated by the researcher, and random allocation is not possible, so causal claims are weaker.
• A pilot study is a small-scale trial of the procedure before the main study. It checks instructions, timing, materials, operationalisation and observational categories, and can reveal ceiling or floor effects. This saves time and money and lets the design be refined.
• Exam tip: when asked to design a study, state the design, give the operationalised IV and DV, name one control and justify it, and say what a pilot would check.

In more detail

• Worked design: 'Does background music affect revision?' → IV: music vs silence; DV: words recalled out of 20 after 10 minutes; design: independent groups, so the same word list can be used without practice effects; controls: same room, time of day, list and instructions.
• Choosing direction: use a directional hypothesis if past research (e.g. Grant et al.) suggests which way the result will go; otherwise use a non-directional one.
• Worked matched pairs: pair participants on a memory pre-test (e.g. both scored 14 out of 20), then randomly allocate one of each pair to each condition.
• Exam tip: every design choice needs a reason linked to the scenario, e.g. 'repeated measures, because memory varies a lot between people'.