Unit 2

• Aims and variables

  aim- identifies the purpose of the investigation. IV- directly manipulated. DV- variable that's measured. Co-variables- variable in a correlation analysis that varies to other variables. Operationalisation- clearly defines what is tested. Confounding- affects some participants and has negative consequences. Extraneous- variables that aren't measured but affect results.
• Hypotheses


  Alternative- hypothesis that predicts the effect of the IV on DV. Directional- predicts the direction of the results. Non-directional- doesn't state a direction of the results. Null- a statement of no affect.
• Data types and sources


  Quantitative- numerical data, it improves validity but is hard to analyse. Qualitative- non-numerical data, gains in-depth data but is subjective. Primary- gathered first hand by researcher (Watson&Rayner). Secondary- already gathered by another researcher (government statistics).
• Research methodologies

  Lab- takes place in an artificial setting, variables are controlled but behaviour isn't natural. Field- takes place in a natural setting, doesn't produce artificial behaviour but hard to control. Natural- the Iv's vary naturally, allows hard topics to be studied but hard to control. Quasi- not an experiment, controlled conditions but its not random.
• Research methodologies

  Participant- observer doesn't manipulate study, behaviour is natural but there can be observer bias. Non-participant- observer isn't involved in study, more accurate observations but observer bias. Questionnaires- a list of questions, large samples can be collected but social desirability bias and response rates may be low.
• Research methodologies

  Structured interviews- questions used to collect data, easy to replicate but not flexible. Semi-structured interviews- discussion in interviews, gets qualitative data but interviewer bias may occur. Self-report- participants report their own thoughts, shows why people behave in certain ways but social desirability bias.
• Research methodologies

  Correlational studies- looks at the relationship between co-variables, shows variable strength but no cause and effect relationship. Case studies- detailed study, can gain in depth data but relies on memory. Content analysis- accesses qualitative/secondary sources, high ecological validity but observer bias may occur.
• Experimental design

  Independent- different participants in different conditions, high validity but no cause and effect relationship. Repeated- same participant in different conditions, eliminates individual differences but reduces validity. Matched pairs- groups are matched, improves validity but is very time-consuming.
• Research locations

  Lab- uses controlled artificial conditions, high control over variables but not a natural setting. Field- natural setting/environment, high ecological validity but lower control over variables and lack of consent. Online- surveys shared on social media, cost effective and reaches large group but social desirability bias.
• Participants

  Sample- size of a sample depends on factors, the sample shouldn't be too large. Sampling frame- a list of those within a target population who can be practically sampled. Target population- a group the researcher is interested in (selective)
• Sampling methods

  Random- equal chance of being chosen but unrepresentative. Opportunity- people who available at that time, quick but small selection. Systematic- system free of researcher bias but is unrepresentative. Stratified- selected randomly from a group, wide population but time-consuming.
• Sampling methods

  Quota- researcher chooses a group, wide population but volunteer bias.Self-selected- volunteers, participants are willing to take part but volunteer bias. Snowball- first participant chooses the next, study hard to access groups but doesn't represent wide population.
• Observational sampling

  Time- very detailed but is time-consuming. Event- more time efficient but limits the level of detail.
• Descriptive statistics

  Mean- calculating the average, it considers all values but can be affected by extreme results. Median- numbers from lowest to highest then the middle, not affected by extreme results but doesnt consider all values. Mode- most common value, good for data in catagories but can be several modes.
• Descriptive statistics

  Range- difference between highest and lowest, easy to calculate but can be affected by extreme values. Standard deviation- how the scores are different from mean, value is affected but time consuming
• Levels of measurement

  Nominal- data is put in separate categories, the simplest form of data(height). Ordinal- measurements are put from lowest to highest(scale of 1-5).
• Levels of measurement

  Interval- the intervals are known and equal, can be negative(temperature). Ratio- there is a true zero point and equal intervals between data(time in seconds)
• Graphical Representation

  Frequency table- number of scores in a given category, data needs to be categorised. Line graphs- plot the relationship between two variables, change in variables over time. Histograms- frequencies of results when data is continuous.
• Graphical representation

  Bar charts- frequencies of results across different categories or conditions. Pie charts- relative sizes of different categories(percentage). Scatter graph- comparing two sets of data, whether there's a connection or not.
• Validity


  Internal- researcher bias, social and observer bias, demand characteristics. Demand characteristics- participants guess the true aims/hypothesis and change their behaviour.
• Validity

  Social bias- participants give answers they think others would want. External- unrepresentative sample, small sample size. Ecological- artificial situation and it doesn't reflect real life.
• Reliability

  Internal- procedures aren't standardised, weather changes and different observers. External- hard to obtain samples, case study method with unique behaviour. Standardised procedures- different participants have different experiences in a study, likely to happen in poor lab experiments.
• Ethical issues


  Confidentiality- ensuring personal data is protected and anonymous. Deception- participants are given full information to give valid consent. Harm- researchers may not be able to predict possible harm to participants.
• Ethical issues

  Values/beliefs- researcher must consider relative risk to participants. Valid consent- participants should be fully informed to give consent. Children- under 16s cant give valid consent. Animals- reduces animals in research, replace animals with alternative and refine procedures, minimalizing harm.
• Dealing with ethical issues

  Ethics committees- ensures ethical guidelines are being followed. Pseudonyms- when valid consent cant be gained a group of similar people are asked if they would consent. Ethical guidelines- respect, competence, integrity and responsibility. Debriefing- informs participants of the true nature of a study.