CHAP 2 / SLIDE 2 Flashcards
5 Steps to Scientific method
- Ask the question
- Design the study
- Collect the Data
- Analyze the data
- Report the results
Some goals of the scientific method are, especially when sampling
- Gather a relatively large , representative sample through random selection to ensure generalizability
- Identify all variables (i.e., confounding variables)
Choose appropriate measures
All 4 of the research designs used in scientific research are:
- Descriptive: “Quantifies” (ie; How common is alcohol use disorder (AUD))
- Case Studies: “In depth examination of the life of one person”
(ie; What does Sam’s AUD look like and how did it develop?) - Observational / Correlational: “Studies naturally occurring relationships (no manipulation)” (ie: “Is age related to risk of alcohol use disorder”)
- Experimental: “Studies impact of manipulation that is randomly assigned”
(ie; Do individuals randomly assigned to group or individual therapy having better outcomes)
2 Types of Variables in Scientific Research
Dependent and Independent Variables
A Quantitative / Qualitative quality that differs and can take different values for different people is known as a _______
Variable
the variable that is influenced by the another type of variable (ie. what we measure)
Dependent variable
a variable that influences another variable
Independent variable
Changing, withdrawing or doing something that doesn’t happen naturally
Manipulation
Studying behaviours in “real world” settings as they naturally occur with no attempt to manipulate
Naturalistic Observation
Advantages of Naturalistic Observation
- We can understand a range of behaviours as they naturally occur
- We can measure “true” behaviours
- Avoids social desirability bias
- High in external validity
Disadvantages of Naturalistic Observation
- Slow progress
- Low in internal validity
- Dosent work well if people are aware they are being Observed
- In-depth examination of the life of One person
- Interviewing the person or people that know them, analysing writings, etc
Case Studies
Advantages of Case Studies
- Find out about personality in great detail
- Formulate a general hypothesis that can be tested on a larger sample
- In depth knowledge about rare phenomena
Disadvantages of Case Studies
-Results cannot be generalised
-Cannot establish causality
- Determining whether there is a relationship between two variables, without manipulation
- Examines the relationship between variables as they occur naturally-
- Correlation Coefficient (+1 -to -1)
Correlational Studies
In a study of happiness, if as:
As self- esteem increases, happiness also increases
&
As self-esteem decreases, happiness also decreases
This would be known as a ________________
Positive Relationship
In a study of happiness, if as:
As self- esteem increases, happiness decreases
&
As self-esteem decreases, happiness increases
This would be known as a ____________
Negative Relationship
The “strength” of a correlational relationship (ie; 0.01 vs 1) is known as the ___
Magnitude
A correlation of +1 or -1 is a ______ correlation
Perfect
In labelling the magnitude of correlations and their Interpretations
0.5 = ___
> 0.5 - 1 = _____
< 0.5 0.1 = _____
Moderate relationship (0.5)
Very Strong - To - Perfect Relationship
Weak - To - Very Weak Relationship
Advantages of Correlational Studies
- Examines the relationship between variables as they occur naturally ( No manipulation involved)
In class example of where a correlational study could be used where a experimental one could not
Ie. To study the role of nutrition in intelligence, we have to manipulate the IV to see an effect on the DV. We now need two sample groups, one of a healthy diet, one group starves, this study is unethical (Impossible for a experimental, but allows opportunity for Correlational as there are individuals that sadly are in the category naturally)
Disadvantages of Correlational Studies
Not designed to identify causal relationships (Directionality problem; third variable problem)
Correlation does not indicate _______
causation
2 Key Requirements of Experimental Methods are
- Random assignment of participants to condition
- Manipulation of an independent variable
2 Main Classes of Experimental Design
- Between-group design
- Within-group design
Apply different interventions to different groups of people and compare the effects on the outcome measure across groups
Between-group Design
Apply different interventions to the same people at different times and compare the effects on the outcome measure owing each intervention
Within-group Design
The Effect of an individuals expectation of whatever they are taking or doing
Placebo Effect
Being exposed to something first, compared to another thing first could have adverse effects in a research study
The Order Effect
Neither Researchers nor participants are aware of whose in the experimental or control group
Double Blind Study
Being unaware whether one is in the experimental or control group
Blind Study
the probability of obtaining our finding merely by chance vs hypothesis (hopefully) small
Statistical Significance
If a difference is observed in the DV - this can either be attributed to ______ or the _________
chance alone
independent variable
What to use to estimate the probability of getting a difference of the size observed merely by chance and its value / estimate statistical significance
P-Value (Probability Value)
P< 0.05
_______________ is an example of individually randomised between-group design
Randomised controlled trial (RCT)
when researchers expose participants to each condition or treatment in different orders to prevent order effect
Counterbalancing
Having researchers aware of study conditions is a potentially important source of __________ bias
Confirmation Bias
a researcher’s expectations about the findings of their research are inadvertently conveyed to participants and influence their responses
Experimenter expectancy effect / Rosenthal effect
Participants pick up cues from studys that allows them to guess researchers hypothesis, and can alter their behaviours
Demand Characteristics (also Experimenter expectancy effect / Rosenthal effect)
- Assuming that characteristics is term for specific thing rather than overall effect
Reviews all research proposals carefully before approval, Including:
* Informed consent
* Duration of the study
* Potential risks and discomfort
* Volunteer participation & right to withdraw
Research Ethics Board
Having a strong reason to believe is different from having ____________________
direct evidence to support an intervention’s effectiveness
How can results from sampling in research be most generalizable?
Large Sample Size (high power)
* Representatitve sample (both sexes, age range, different culture, etc)
What outcome measures?
* To measure abstinence self-reports? Biochemical tests at follow-up? AKA
* Follow “best practice” in the field:
* Alcohol → self-reports
* Smoking → biochemical tests
* Illicit drug use → self-report, but urine screening is essential