Please enable JavaScript.
Coggle requires JavaScript to display documents.
Evaluation PRP (Piliavin) - Coggle Diagram
Evaluation PRP (Piliavin)
RELIABILITY
Reliability reduced bc data collected came from unbalanced trials in each victim condition
Black Cane trials: 16
White Cane trials: 54
Results affected by different statistical weight for each condition
Field Experiment, however still high level of control bc of standardisation
Clothes of Victim: Eisenhower jacket, old slacks
Time of Victim Collapse: 70s
Model Interventions: 70s or 150s
Duration Exp: 7 1/2 min
Ensures that study is easily replicable, therefore High Internal Reliability
Event Sampling (short periods of bystander behaviour observation) makes it easier for data to be accurately/ reliably recorded. This is bc data = objective
Ex: No of bystanders in critical area
Race of helpers
Difficult to misinterpret. Allows for easy comparison between replicated studies and og PRP study
VALIDITY
Low control bc natural environment (field exp) which may affect behaviour of participants
Ex: participants may have viewed exp more than once
This means that participants would not have exhibited natural behaviour to the emergency making for Low Internal Validity
High Internal Validity bc procedure = single blind
Participants unaware behaviour is recorded (observation) therefore likely for behaviour to be natural. Makes for no demand characteristics. Deception prevents demands characteristics and therefore protects the Internal Validity
Low Population Validity bc opportunity sample used
Only ppl that traveled by metro or happened to be there included, unrepresentative of target pop (not everyone has equal chance of inclusion bc not everyone uses metro)
Ppl traveling by metro = may share specific characteristics, making findings un-generalisable to target pop
Facile Generalisation- large sample size
4450 men & women
45% black, 55% white (representative of NY pop)
Therefore findings will be more easily generalised to target pop compared to studies w smaller samples
Low Ecological Validity- Decreased genralisability bc exp took place in a metro, a confined space. Participants were captive for 7 1/2 min w victim, can't escape.
Real- life situations seldomly include inability of escaping, therefore inapplicable to most real- life situations
High Ecological Validity- field exp carried in real life environment
Ny subway with real individuals, emergency is also highly realistic (mundane realism of participant task)
Because of this, findings can easily be applied to real life situations.
ETHICS
Deception- Participants deceived abt participation, aim & data collected. They believed the exp was a real life situation.
This could: cause guilt distress momentarily exposing them to harm
Also: inability to give free informed consent
However: Deception necessary to limit demand characteristics (for observers to observe natural participant behaviour) and therefore to protect Internal Validity
Protection - responsibility to protect participant from any harm greater than that encountered in daily life.
Ex: many participants left critical area indicating distress
Also: since opportunity sample, researchers could not later contact participants for debriefing. This means they cant be sure harm = not long lasting.
However; Help offered in most situations (Model used 3 times) indicating situations did not severely distress participants
Further: Distress only temporary & decreases when help offered.
Participants also able to leave critical area if uncomfortable
Right to Withdraw- Single blind, participants unaware they were in a study, thus they could not freely consent.
Opportunity sample- participants could not quit at any time or withdraw their results from study as they did not know they were part of it in the 1st place.
Explanation
Reductionism: Not Reductionist. Exp measures takes into consideration multiple factors possibly influencing helping behaviour. Ex: Gender, Race, Size of Group, Victim conditions, presence of model. Shows: researchers aware that helping behaviour is complex, several victim characteristics necessary to understand why ppl do/ don't help
Individual: Cost- Reward Analysis of helping is made by each individual. It's influenced by Physical/ Psychological traits of each individual.
Ex: 90% of helpers were male, slight tendency towards same race helping.
Shows: helping behaviour is affected by characteristics of victim/ bystander alone
Situational: External factors involved in helping behaviour considered
Ex: No of bystanders, Race of victim, condition of victim, influence of model
Diffusion of Responsibility unsupported bc of proximity to victim