Experimental Method, Aims and Hypotheses

Experimental Method

• Concerns manipulation of the independent variable (IV)
• Determines its effect on the dependent variable (DV)
• Experiments can be in a field, laboratory, or can be quasi or natural experiments

Aims

• General statement by a researcher
• Indicates what the researcher plans to investigate and the purpose of their study
• Aims come from theories and reading similar research

Hypotheses

• Precise statement expressing the relationship between variables
• Can be directional or non-directional
• Directional hypotheses state the direction of the relationship
• Non-directional hypotheses do not state a direction
• Example of a directional hypothesis: "The more sleep a participant has, the better their memory performance"
• Example of a non-directional hypothesis: "The difference in sleep hours will affect memory performance."
• Directional hypotheses suit studies with existing research suggesting a specific outcome
• Non-directional hypotheses suit studies lacking previous research or with contradictory research

Independent Variables

• Manipulated by the researcher to affect the DV

Dependent Variables

• Measured by the researcher
• Affected by a change to the IV
• Extraneous variables need controlling to ensure confident conclusions on IV to DV effect

Variables

• Operationalisation involves the act of clearly defining variables in terms of measurement
• Variables should be defined and measurable
• Hypotheses should show operationalisation

Variables Example

• A better operationalised directional hypothesis: "Participants who get at least four hours of sleep will perform better on a memory test, achieving higher scores than participants who get less than four hours of sleep"

Control of Variables

• Experiments work best when the only thing that affects the DV is the IV
• Variables that interfere with the IV or DV need to be removed/ controlled
• Confounding and extraneous variables should be dealt with

Extraneous Variable

• Nuisance variables
• Affect the DV
• Don't vary systematically with the IV
• Examples include lighting in the lab and age of participants
• Make it harder to detect effects

Confounding Variable

• Other variables that affect the DV
• Unlike extraneous variables, confounding variables change systematically with the IV
• Obscure the origin of the DV, as the confounding variable could be the cause
• Example: time of day when the experimental task is done can effect tiredness of participants
• Potential confounding variables need to be identified and controlled such as the environment

Demand Characteristics

• Cues that researchers give
• Makes the participant feel like they can guess the aim of the investigation
• Can cause participants to act differently within the research situation, known as participant reactivity
• Participants might change their behavior to fit the situation rather than naturally, which affects validity • Participants may try to act to fill the researcher's expectations called the 'Please-U effect' • Some perform deliberately badly to ruin the study, this is known as the 'screw-U effect'

Investigator Effects

• Unwanted influence from the researcher's behavior on the DV
• Includes the study design, participant selection, and interaction with each participant

Randomisation & Standardisation

• Methods used to minimise the effect of extraneous/confounding variables

Randomisation

• The use of chance reduces bias from investigator effects
• Materials, order of conditions, and participant selection can use this

Standardisation

• Using the same procedures and instructions for each participant
• Eliminates non-standardised instructions as extraneous variables

Types Of Experiments

Laboratory Experiment

• Takes place in a controlled environment
• Variables can be carefully controlled
• High degree of control, experimenters control all variables
• The IV is replicated precisely, leading to greater accuracy
• Researchers can repeat experiments and check results
• Artificial situations can produce bias
• Has low ecological validity
• There is a risk of Experimenter Bias
• Participants may be influenced by expectations

Field Experiment

• Conduct experiment in a natural environment, but variables are still controlled
• Naturalistic, increases the validity of the experiment
• Higher ecological validity, but there are still ethical considerations to be aware of
• More difficult due to the lack of variable precision

Quasi Experiment

• The IV is not determined by the researcher
• Can be easily replicated
• There may be confounding variables

Natural Experiment

• IV not brought on by the researcher
• Would happen regardles
• Opportunities for research that would otherwise be impossible arise due to practical or ethical reasons
• It can be hard to randomise

Sampling

• Researchers need to choose the best way to select participants to take part in their investigation
• The Population is a group of people from whom the sample is drawn

Opportunity Sampling

• Participants have to be available at the time of the study
• They get recruited conveniently due to the ease of access

Random Sampling

• Each member of the population has equal chances of being selected
• Each member of the population is assigned a number
• Numbers are randomly picked such as the lottery system
• There is less researcher involvement

Systematic Sampling

• A predetermined system applied consistently decides who gets picked
• A system is used selects every nth member from the sampling frame

Stratified Sampling

• The sample reflects the proportions of people in subgroups within the wider population
• Identify subgroups, calculate the required proportion from those
• Select randomly from each

Volunteer Sampling

• Self selection
• Participant offers to take part in response to an advert
• Quick access to willing and helpful participants
• Easy to conduct

Experimental Design: Types

Independent Groups Design

• Participants only perform in one condition of the IV
• No order effects from practice/ boredom
• Participants less likely to guess the study aims
• Less control over the participant variables

Repeated Measures Design

• Participants take part in all conditions of the IV
• Eliminates participant variables
• Fewer participants required
• Not as time consuming finding and using them
• Order effects are presented e.g. boredom may mean in second condition done participant does not do as well on task

Matched Pairs

• Pairs first matched on some variable that has been found to affect the dependent variable (DV)
• One does one condition, another does another
• There are no order effects
• Demand characteristics are less of a problem, but it can be hard to match participants into a pairs

Pilot Studies

• A smaller study (or small-scale version of the investigation) that is undertaken before real implementation
• Looks to prevent likely problems that can be identified
• To modify to deal with problems early
• Allows money and time to be saved

Single and Double Blind Reviews

Single Blind

• Researches withold wether the participants are assigned the treatment procedure or the control procedure
• To ensure that the participants don't alter results due to preconcieved expectations and demand characteristics

Double Blind

• Neither the researchers or participants know who is assigned the treatment procedure

Control Groups

• Set a baseline
• Results from the experimental condition can be compared to see if the IV changes anything

Observational Techniques

Naturalistic

• Watching and recording behaviour in the setting where it would normally take place.
• High ecological and external validity
• Low validity if any of the participants become aware they are being watched

Controlled

• Watching and recording behaviour in a structured environment
• Researcher can more easily focus
• More control over extraneous and confounding variables
• Increased mundane realism
• Easy replication

Overt

• Participants behaviours are recorded and watched with the full understanding and consent

Covert

• Natural and representative behaviours are recorded
• Data has high internel validity due to these conditions of natural behaviour

Participant

• Researcher is observing and is apart of the group

Non-participant

• Researcher observes from a distance

Observational Designs

• Observer bias, when an observer's reports are based on expectations
• To check inter observer reliability of the observation a inter observer reliability test is done: (Total number of agreements / total number of observations) x 100
• A score above 80% determines high inter observer reliability

Unstructured designs

• Continuous recording is done where the researcher writes everything they see during an observetion

Structured

• Quantifying observed aspects using a predetermined list of sampling methods
• Results are easier to work with

Behavioural Categories

• Target behaviour is broken down into smaller more measurable components

Types of samplings

Time Sampling

• Recording behaviour within a timeframe pre established before the study

Event Sampling

• Involves the counting of particular behavioural events as carried out by a target

Correlations

• Mathematical method used to investigate an association

Correlation Vs. Experiments

• Covariables are looked at, instead of maniluated
• There is no found cause and effect

Three correlation graphs and their properties

Negative Correlation

• While one veriables increases, the other decreases

Positive Correlation

• While one veriable increases, the other also increases

Zero Correlation

• There is found no relationship between veriables

Further Data types

Curvilinear Relationship

• As one cariable increases, so does the other only to a certain point

Hypthesis

Directional Hypothesis

• States wether there is a negative, or positive correlation

Non Directional Hypothesis

• Only points to correlation but does not specify the type

Correlations: Strengths and Weaknesses

Strengths

• Used as starting points to assess patterns between covariables

Weaknesses

• Hard to extablish the cause and effect

Datatypes:

Qualitive

• Displaed in words
• NON-numerical

Quanititave

• Displayed Numerically
• Graphs available

Primary data

• Data is obtained by the researcher
• Targets exact information needed

Secondary data

• Data obtained by someone other than the researcher
• Requires no minimal effort to collect

Meta-analysis

• Data is studied as a large ammount
• Higher generalisability

Measures of central tendency

Calculations

• Arithmetic average
• Total of all values devided by variables

Measures of dispersion

Presentation and display of Quanitative Data

Tables

Properties

• Data can be organised and descriptive statistics are used

Bar Charts

• Used for discrete data, which describes data that has been divided into categories.
• Bars do not touch

Histograms

• Bars touch
• This represents continuous data.
• The x-axis has equal sized intervals of one category,

scattergrams :

Properties

• Shows association between covariables
• Either of the covariables can occupy either axis, and each point signifies both x and y

Distributions

Normal distribution

• Symetrical patter of frequency data creating a bell

Skewed Distribution

• Spread of frequentc data that is not sysmetric
• Data clusters to one end

Peer Review

AO1

• An assesment of scientific work by experts

Why its is done

• To make sure research is of a high level

Main Purposes

• To tell why the research is worthwhile
• To confirm the relevance and quality
• To suggest improvements

AO3 - problems

• Anonymity a is problem
• Publication bias may be present
• It can be dificult to find an expert

Economic Implications

• Research shows how psychology research supports the countries economic prosperity
• Absence due to stress costs billions
• These problems help find why research is important

Social influence examples

• Social influence leading to social changes

Memory example

• Eyewitness testimoney

Case studies

Propertys

• Detailed _Qualitive- data, life expereinces

Stengths

• Gain in depth insight

Limitations

• Data gathered from one person

Bias

• Socially disireable bias, interpretative bias

Contecy analysis

Propertys

• Studying human behaviour indirectly

Levels of measurment

Nominal data

• Data is classified by its name or category

Ordinal data

• Data is presented in an order
• No equal intervals

Interval data

• Data is calculated by numerical scales

Method in science

Abstact

• Key summary of details
• 150_200 words

Introduction

• Mentions prior research

Results

• Shows study findings with inferential statistics/ thematic analysis

Testing

Statistical testing

• A way of determining if a hypothese should be rejected/or accepted

Sign test- 3 properties

• Looked for a difference, not an association
• Used a related experimental design
• Repeated measures design
• Collected nominal data

How to conduct

List Hypothesis

• Includes both alternative and null

Record Data

• Record data and work out the sign. For example, the sign will be negative if the value has decreased

The Design

• Whether it involved an unrelated design

Level Of The Data

• The data either uses ordinal, nominal, or interval

Ethical issues and resolutions

• The ethical issues that occur must first be identified

• A cost benefit analysis might be helpful early on to decide if the study is worth undertaking

  • Types of ethical problems

    • Informed consent
    • Protection from Harm
    • Privacy and Confidentiality

Self Port Techniques and Design

• Self port helps researchers gain information from their subject in a non esperimetal environment
• Can be delivered in several ways
• Questionnaires
• Interviews

Qusetionairs

Propertys

• Helps the researcher assesses subject based off questions in a written manner

Open & Closed Questions

• Open - No restrictions, free to answer but can be difficult to convert to statistical data easier and more concise
• Closed - Only answer though the responses, can be very frustrating and difficult for participants

Advantages

• Cost Effective
• Efficient

Disadvantages

• Difficult for reseracher to know if the target polpulatopm filled it out

Creating Quetinnares

• Must remain clear, concise, and easy to understand
• Must avoid jargon or un ecucated languege

Interviews

Propertys

• Live conversation between interviewer and subject
• More free flowing

Advantages

• Higher standarisation

Disadvantages

• Interviewer bias

Bias

• Neutral questions
• Ethical qustions

Reliability

Types of reliability

• Internal Reliability
• External rediability

The way to assess each type, respectively are

• Split half method
• Test retesr

Ways to improve reliability

• Re written Quesionaires
• Similar to questionnaire