Fichas
Psyc
Fichas del conjunto (23)
- 1Three research designsExperimental
- independent groups
- repeated measures
Quasi-experimental
- not quite an experiment
Correlational - 2Independent group designtrue experiment (one that you can be confident that the results are due to what you were testing) and relies on random division of people into groups.
- 3Why an experiment?Random assignment of participants to group A or B eliminates systematic differences between groups (e.g. IQ, motivation, political affiliation.)
- 4Independent Variabledirectly manipulated by the experimenter
- 5dependent variableis presumably affected by the IV because this is the only systematic difference between the groups.
- 6the control group (one that doesn't change behavior) is the..Baseline for comparison
- 7Why do an experiment?Casual evidence about the relationship between two things.
CAUSAL RELATIONSHIPS ARE THE MAIN ADVANTAGE! - 8Disadvantage of experimentrandom allocation is impossible at times.
when comparing groups with specific traits like age, IQ and visual acuity we cannot manipulate
ethical concerns - would people actually want to participate in the experiment.
Do you really want to test the effects of alcohol on yourself? - 9Quasi-experiment!!!find people who already fit the criteria!!!
Since you cannot randomly assign you try make them as similar as possible, age, gender, how long they spend in the gym each day. Although sometimes the reason people already do the activity like reading self help books is because they already have mental problems. - 10Correlational studyfind 100 people and measure how many minutes per day they spend reading self-help books.
Measure how often they visit a Psyc each year.
Record data on a scatter plot and use stats to assess whether there is a relationship between minutes per day and how many times they see a psyc. - 11Quasi-experiments and correlational studies.They do not allow us to make causal inferences.
- 12Advantages of Quasi-experiments and correlational studies.can be used when random allocation to groups is unfeasible
Naturalistic - ecological validity (the participants are assigned on their actual real world behaviors) - 13Disadvantages of Quasi-experiments and correlational studies.cannot infer causality. if there is a relationship between self-help reading and psychologist visits we cannot conclude that reading self-help books causes fewer visits to a psychologist.
- 14Good experiments gone bad? Random variability:The name given to differences or variability in the dependent variable not attributable to the independent variable. (like when taking a test sleep and other situational factors can affect the results)
- 15three main sources of random variability;individual differences; IQ, nature/nurture, some people are just better suited to study.
Situational variable; Noise
Measurement Error; A test doesn't ask everything so its not 100% reflective of the students knowledge (you will always have measurement error its incredibly hard to have a perfect test. - 16Extraneous and confounding variablesindividual differences and situational variables are extraneous variables
- they make the IV harder to detect
When extraneous variables change systematically with levels of your independent variable, you can't be sure which one is causing changes to the dependent variable.
for example if you found 50 men that read self help books and 50 women who just read normal books we cannot tell if the self help books are the cause because one whole group is make and one whole group is female. - 17Prevent bad experiments* eliminate situational and confounding variables.
- so you can interpret the relationships between the independent an dependent variables
* Nullify/minimize individual differences.
to improve your changes of finding a relationship between the Independent variable and dependent variable, if one exists. - 18How to reduce situational variabilityHold such variables constant as much as possible.
- test in quite room
- give standard instructions
- use recorded instructions
- read from a script
- use the same experimenter
Balance variables across conditions
- time of day (25 people in one group in the morning x two groups, 25 people per group in the afternoon x two groups.)
- order of conditions
Match on relevant factors
example: age (if you want to measure Dr's visits age would matter. (85 and 25 go different amounts)) , gender, visual acuity, health.
Pre-screen for these (i will screen for physical health whilst measuring mental health as people with terminal diseases might have worse mental health) and/or pre-test/post test design (measure everything again to see if they have changes after experiment).
Use more participants
on average, more people = the differences between participants will balance each other out. - 19Repeated measures designPick 50 random people and make them read self help books. See how many times they go to the psyc. then the next year read normal books and see how many times they go to a psych. TEST THE SAME PEOPLE.
Each participant takes part in every condition
holds random variability due to extraneous variables constant across all conditions.
Differences between conditions can be attributed to the IV. Its just people compared to themselves.
another good thing is that it halves the number of participants needed. - 20Repeated measures design downsidescan introduce confouding variables
- Practice; if they have done it before they have practiced. They might be better as they have already done it.
- fatigue, participants may be bored and tired.
- carry over (say you are testing caffeine you give condition A caffeine (or drugs) and you go right to condition B and they might still have caffeine in them.
- sensitization, when doing an experiment you may work out what the experimenter is looking for and act differently. - 21Repeated measures design downsides solutionscounterbalancing or randomize condition order. half of the people will do condition A first and some people will do condition B first.
randomize condition order is when you have a complicated design so each group is randomly designated a start condition. Random orders of conditions, not as good as counterbalancing but will still help balance effects. - 22describing datausing FIGURES to visualize how 'most people do' (central tendency) and calculations of variability.
Calculation of central tendency
- mean, mode, median.
calculations of variability.
GRAPHS AND VISUALS ARE IMPORTANT IN HELPING OTHER PEOPLE TO UNDERSTAND AND YOU TO UNDERSTAND. - 23frequency tablessimple fre
Información
- Descripción
- Fichas sobre Psyc , creado por Samantha Hopson el 06/04/2018.
- Creado por
- Samantha Hopson
- Visitas
- 0
- Valoración