Science practises - Research methods and data analysis

Empirical approach

An evidence-based method that draws on observation and experimentation.

Scientific Attitude

Curiosity - does this work?

Skepticism - what do you mean and how do you know?

Humility - explore further after anomalies.

Critical thinking

Thinking that does not blindly accept arguments and conclusions. Rather, it examines assumptions, discerns hidden values, evaluates evidence, and assesses conclusions.

- Examine everyone's assumptions in-depth

- Discover hidden biases

- Evaluate evidence and its source

- Assess conclusions logically and in context of data

Hindsight bias

Also called the "I knew it all along" effect

The tendency to believe, after learning an outcome, that one would have foreseen it

Overconfidence

The tendency to be more confident than correct—to overestimate the accuracy of our beliefs and judgments.

Theory

An explanation using an integrated set of principles that organizes observations and predicts behaviors or events

- Conclusions cannot be drawn before they

- Good theories produce testable predictions and hypotheses

Hypothesis

A testable prediction, often implied by a theory

Can support or disconfirm theories

Operational definition

A carefully worded statement of the exact procedures used in a research study

- Allows for replication

- Define terms more clearly (what is considered this?)

- Reduce bias

Replication

Repeating the essence of a research study, usually with different participants in different situations, to see whether the basic finding extends to other participants and circumstances

Case studies

A descriptive technique in which one individual or group is studied in depth in the hope of revealing universal principles

Case studies only lead to ideas, so other methods are needed to follow up

Naturalistic observation

A descriptive technique of observing and recording behavior in naturally occurring situations without trying to manipulate and control the situation

Mostly used for recording data

Surveys

A descriptive technique for obtaining the self-reported attitudes or behaviors of a particular group, usually by questioning a representative, random sample of the group

- Requires a representative population

- Need to be aware of wording bias and sampling errors

- People can be untruthful when questioned about themselves

Sampling bias

A flawed sampling process that produces an unrepresentative sample

To reduce it, wording must be carefully chosen to avoid negative connotations and a random sample must be chosen from a representative population

Random sample/selection/assignment

Fairly represents a population, because each member has an equal chance of inclusion

Population

Refers to all those in a group being studied, from which random samples may be drawn. This does not refer to a country's whole population

Correlation

A measure of the extent to which two factors vary together, and thus of how well either factor predicts the other.

Variables

Any factors that can vary, and are feasible and ethical to measure

Correlational study

A descriptive study that looks for a consistent relationship between two variables

THEY DO NOT SHOW CAUSE AND EFFECT RELATIONSHIPS

Correlational coefficient

Statistical index of the relationship between 2 variables (from -1 to +1)

Positive means variables move in the same direction

Negative means variables move in inverse/opposite directions

Note: The words "positive" and "negative" do NOT mean good or bad in correlations

Scatterplot

A graphed cluster of dots, each of which represents the values of two variables

Slope suggests direction and strength of relationships between variables

Amount of scatter suggests the strength of the correlation, with the less scatter, the stronger the relationship

Illusory correlation

Perception of a relationship where none exists, or perception of a stronger relationship than actually exists

Feed an illusion of control; falsely instilling confidence in the actions we believe will result in favourable outcomes

Regression towards the mean

The tendency for extreme or unusual scores to fall back (regress) toward their average.

Extreme results tend to be caused by unusual combinations of circumstances, and they tend not to repeat, so extreme behaviour tends to regress towards the average

Failure to recognise regression towards the mean can create superstitious thinking, mistaking the influences on changes in behaviours

Experiment

Research method where an investigator manipulates one or more factors (independent variables) and observes the effect on a certain behaviour (dependent variables) to identify a cause-and-effect relationship between the variables

1 or more variables are manipulated while the rest are held constant in order to isolate effects

Experimental group

The group exposed to the treatment - one version of the independent variable

Control group

Group not exposed to the treatment; contrasts with experimental group and serves as a comparison for evaluating the effects of the treatment

Random assignment

Assigns participants to experimental and control groups by chance, thus minimising preexisting differences between each group

Single-blind procedure

Experimental procedure in which the research participants are ignorant(blind) about whether they have received the treatment or placebo

Placebo effect

Used to describe experimental results caused by expectations alone; any effect on behaviour caused by the administration of an intert substance, which the recipient assumes it is an active agent

Independent variable

The factor that is manipulated by the researcher to determine its effect on another variable

Must be operationally defined

Confounding variable

A factor other than the independent variable that might influence a study's results in an experiment

Controlled by randomly assigning participants to the experimental and control groups

Dependent variable

The outcome factor; the variable that may change in response to manipulations of the independent variable.

Must be operationally defined

Validity

Accuracy -The outcome that is measured; extent to which a test or experiment measures or predicts what it is supposed to

Random assignment and other features in research design can help to establish more valid findings

Reliability

Consistency - can be achieved through replication which is fostered by creating operation definitions in experimental designs

How to determine which research method to use

Factors:

- Appropriate method based on context

- Cost

- Time available

- Ethical issues and other limitations

Qualitative research

Relies on quantifiable, numerical data

Informed consent

Giving potential participants enough information about a study to enable them to choose whether they wish to participate

Debriefing

The post-experimental explanation of a study, including its purpose and any deceptions, to its participants

Research Rules

- Researchers must obtain informed consent, giving participant enough information to make a choice to participate

- Harm and discomfort must be limited

- Participants' information must be kept confidential

- Participants must be fully debriefed and given details about the purpose of the research, its findings, and deception which may have occured

- Institutional review boards are set up to review and approve research proposals

Descriptive Statistics

Numerical data used to measure and describe characteristics of groups

- Includes measures of central tendency and measures of variation

- Help researchers measure and describe how a data set about a particular group might look

- Can often show patterns in data but do not help a researcher draw causation

Histogram

Bar graph depicting frequency distribution

- Show frequency distribution

- Y-axis may be manipulated to make differences look bigger than the data truly represents

Be a critical thinker: consider scale labels and ranges of histograms

Mode

Most frequent occuring score in a distribution

Mean

The arithmetic average of a distribution, obtained by adding the scores and then dividing by the number of scores

Median

The middle score in a distribution; half the scores are above it and half are below it

Percentile rank

The percentage of scores below a specific score in a distribution of scores

Range

Describes the difference between the highest and lowest scores in a distribution

Skewed distribution

Describes a representation of scores that lack symmetry around their average values

- The mean's validity is most affected by extreme scores which can created a skewed distribution

- Median score is not affected this way

Standard Deviation

Computed measure of how much scores vary around the median score

Normal curve

Symmetrical, bell-shaped curve that represents the distribution of many types of data

- Most scores fall near the mean (68% per deviation which is 34% on both sides, 95% per 2 deviations, which is 47.5% on both sides)

Inferential statistics

Numerical data that allow one to generalise; to infer from sample data the probability of something being true of a population

Distinguishes chance from meaningfulness

Meta analysis

Statistical procedure for analysing the results of multiple studies to reach an overall conclusion

Combining findings of multiple studies enhances validity

3 basic ideas to generalisation

- Representative samples are better than biased or unrepresentative samples

- Bigger samples are better than smaller ones

- More estimates are better than fewer estimates

Statistical significance

Statistical statement of the likelihood that a result occurred by chance, assuming there is no difference between populations studied

If difference between means are large, then a real difference in the population is more likely to exist

Probability/P-values are used to represent statistical significance and must be lower than 5% (0.05)