Research Methods

Module 1:

Scientific Method

  • Used to ensure reliable and meaningful research 

  • A standardized process to minimize biases, conflicts and oversights

Scientific Method Steps
  1. Theory

  • Theory - A general set of ideas about how the world works

  • Research past works 

  1. Hypothesis

  • Hypothesis - Testable statements making specific predictions about relationships between variables

  1. Research Methods

  • How the hypothesis is tested

  1. Collect Data

  • Measure outcomes

  1. Analyze Data

  • Discover trends/relationships between variables

  1. Report Findings

  • Publish articles in scholarly journals

  1. Revise Theories

  • Incorporate new information into our understanding of the world

Paradigm Shift - A particularly dramatic change in our way of thinking

Conducting an Experiment

Anecdotal Evidence - evidence gathered from others/ones’ experiments

  • Hearsay

  • Personal experiences

  • Not sufficient for Science

Experiment - measure the effect of 1 variable on another

Independent Variable - manipulated by the scientist

Dependent Variable - being observed by the scientist

Control Groups
  • Experimental and control groups should be as similar as possible

Experimental Group - the independent variable IS manipulated

Control Group - the independent variable IS NOT manipulated

Within-Participant Design - manipulating the independent variable WITHIN each participant

Between-Participants Design - 1 experimental and 1 control group

Within-Participant Design

Between-Participants Design

Pros

Minimizes the effect of participants' differences

  • Prevents carryover effects

  • Shorter experiment length

Cons

  • Time-consuming

  • Costly

  • Practice effect

  • Larger sample sizes needed

  • More resources needed

  • Confounding variables

Practice Effect - improvement over time(the course of the experiment) from experience

  • Ard to separate from the multiplication of the independent variable

Confounding Variables - Variables that obscure the effects of the independent variable on the dependent variable

  • Any variable related to the independent variable

  • Leads to questionable results

    • Makes it difficult to draw conclusions

Sampling

Sampling tries to eliminate all confounding variables

  • But it is difficult to find people who fit ALL the requirements

Results from very specific groups CAN NOT be generalized to other groups

Population - A general group of people we want to learn about

Sample - subset of individuals selected from the population

Random Assignment - Participants are randomly assigned control or experimental group

Random Sample - A group of participants selected from a larger population where everyone has an equal chance of being chosen

  • Most representative of the real population to avoid biases

  • Happens after random assignment

If both random assignment and sampling are truly random, we are reasonably confident that any pre-existing differences are limited

Avoiding Biases and Placebo Effect

Placebo Effect - An individual shows a response to a treatment that has no therapeutic effects

Participant Bias - Intentionally or unintentionally bias their results to match expected results

Experimental Bias - Intentionally or unintentionally influencing results to favour their hypothesis

Experimenter Bias - actions by the experimenter, intentionally or unintentionally, that influence the results

Blinding - A research method that conceals information to prevent bias in an experiment, ensuring more accurate results

Blind Experiment - Participants don’t know the group they are in or the treatment they are receiving

  • Minimizes participant bias

Double Blind Experiment - neither the experimenter nor the participants know what group they are in

  • Minimizes experimenter bias


Module 2:

Histogram - A type of graph that reports the number of times bins appear in a data set

  • Base for frequency distribution graph

Bins - Groups of values

Frequency Distribution - Illustrating the distribution of how frequently values appear in a data set

  • The line shows how likely something is to occur in the distribution

 

Normal Distribution Characteristics
  • Smooth 

  • Symmetrical

  • Bell-shaped

  • Curve with a single peak

Every day measures are normally distributed

  • E.g. IQ, Height, Test scores

Measures of Central Tendency 

Measures of Central Tendency - where a data set is centred

Mean - the average value of a data set

  • Add all data points, then divide by the number of data points

  • Can be shewed by outliers

    • Outlier - extreme point, distant from other data points

Mode - value that appears most frequently in a set

  • The typical response

  • Can be used for mom-muberical data sets

Median - The center value in a data set when arranged numerically

  • Not shewed to 1 side

  • Only focus on the center/typical value → ignoring other data points

Measure of Variability 

Standard Deviation - A measure of the average distance of each data point from the mean

  • If we only used measures of central tendency, both examples would be the same

Smaller spread = smaller standard deviation

Larger spread = larger standard deviation

Inferential Statistics 

Inferential Statistics - use data sets to make inferences about the overall population

T-Test - uses data points from both experimental and control groups to calculate the probability that samples are from the same population

  • Produces a P-Value

P-Value - probability of the T-Test

  • From 0-1

  • Is the difference large enough?

  • Statistically significant is <0.05

    • Less than 5% that is the difference between the groups, is by chance

Statistical Significance - when the difference between 2 groups is due to some TRUE difference between the properties of the groups, it IS NOT due to random variation

Types of Errors

Type 1: False Alarm - believing there is a difference when a difference DOESN’T exist

Type 2: Miss - Failing to see a difference when a difference DOES exist

Type 1: False Alarm

Type 2: Miss

Examples:

An ineffective drug is believed to be effective

An effective drug is believed to be ineffective

Think:

A man told that he is pregnant

A very pregnant woman is told not pregnant 

Observational Research 

  • Used when it is difficult or impossible for scientists to perform experiments

    • E.g. Ethical or practical concerns

  • Principles of experimental design also apply to observational research 

Correlation - a measure of the strength and direction of the relationship between variables

  • How close are data points to the trend line?

Correlation Coefficient (r) - indicates the strength and direction of correlation

  • from -1 to 1

  • NOT RELATED TO SLOPE

r value 

Correlation

1

Perfect Positive

0.25

Weak Positive

0

None

-0.25

Weak Negative

-1

Perfect Negative

Correlation DOES NOT EQUAL Causation

Strength = the number

  • How far the number is from 0 on either side

Direction = the sign

  • = Positive

  • = Negative