Please enable JavaScript.
Coggle requires JavaScript to display documents.
Lecture 3: Quality indicators of research & research ethics - Coggle…
Lecture 3: Quality indicators of research & research ethics
Ethical guidelines by British Psychological Society (BPS
Valid consent
given by appropriately able and responsible participants
no pressure or coercion
can withdraw without penalty or need to give
reason
issues with valid consent
if participants infer real purpose of research, it might become meaningless as they change behaviours
there is a tightrope to inform while keeping true value hidden
social or sensitive issues with even more care as they might change behaviour
deception is misleading participants about the real nature of study so they do not change behavior
deception should be avoided unless it is essential to research objectives and the research has strong scientific merit
deception should be temporary
deception should be tackled in DEBRIEFING
then this needs appropriate risk management
Limit use of Deception
Debriefing
minimize negative impact of research, especially if deception has been used
participants given full explanation of features and true aims and reasons, and PROFESSIONAL HELP be given if there was distress, etc
participants may exercise their right to withdraw from a study
Minimal risk
must not expose participants to greater physical and psychological danger than their real lives
vulnerable/at-risk people to be avoided
5.Confidentiality
information gained from participants must not be divulged to third parties without consent, unlesss agreed upon in advance. anonymity to be protected
Type of research method
Selected non-experimental methods that yield quantitative data (i.e., survey, observational studies)
inter-coder reliability
degree to which coders of the question or image assign similar answers/same codes
Saturation validity
the collecton and codiing of additional data until no further insights are gained
Experimental methods that yield quantitative data (e.g., laboratory, field, quasi experiments)
Replicability reliability
Internal validity
What is it?
Extent to which the changes in the dependent variable in an experiment really are caused by the manipulation of the independent variable (and not by extraneous variables)
Threats to internal validitty
Demand Characteristics
Characteristics of the experiment that, from the participant’s perspective, seem to ‘demand’ certain kinds of response
Can lead to participants attempting to infer the hypothesis and trying to respond accordingly – perhaps to be a ‘good subject’ and to therefore confirm that presumed hypothesis giving responses that are socially desirable.
How to control for demand characteristics
Single blind study
Information that could introduce bias is kept from the participants, e.g., Participant does not know whether they are a member of the control group or the experimental group
minimize extraneous variables
related to the presentation of the independent variable and the measurement of the dependent variable to minimise inferences about the hypothesis
Eg. instead of The effects of violent television programs on aggressive behavioe, it's understanding human behaviour
Experimenter effect
Ways in which the experimenter may inadvertently convey their expectations to the participants and thereby change the way the participants behave
(Rosenthal, 1976)
Analogues to experimenter effects in questionnaire surveys and qualitative research
researchers might treat participants differently according to their group membership
e.g. 'What do you think of sexes opportunities' might be worded differently to a male and female
How to control
Experimenter does not know whether the participant is a member of the control group or the experimental group. (Single blind as well)
Double blind study
neither the participants nor the experimenter know to which group the participant has been allocated
requires the involvement of a third party who does not know which participants are allocated to which group
external validity
degree to which results are generalisable beyond the setting in which the research is carried out, and have relevance to the real world.
high external validity = mundane realism
Experimental ralism
Extent to which participants take the situation seriously and become involved in the experimental procedure the psychological processes the participant uses in the study can thus reflect those they would use in the relevant real world situation extremely high experimental realism may lead to demand characteristics
All non-experimental methods that yield qualitative data (e.g., surveys, interviews, focus groups)
Test-retest reliability
questions asked more than once in order to see if what they are measuring is in a consistent manner
face validity
whether the questions address what they intend to address
similar to mundane realism
construct validity
whether the questions accurately investigate the phenomenon they want to address
similar to EXPERIMENTAL realims
Replicating Milgram's study (Burger, 2009)
Partial replication Rigorously safeguards the welfare of participants (e.g., a two-step screening process for potential participants). repeated the experiment up to the point at which Milgram’s participants first heard the learner’s protests (150 volts) since 79% of those participants who went past 150volts continued to maximum shock level, can assume similar trends Results: Found similar levels of obedience to Milgram
Research ethics: Ethical principles that govern research have changed over time - Milgram’s precise research design would be unlikely to be ethically acceptable today. theoretical implications (relate to lecture 13 – obedience): the factors that Milgram identified as driving obedience come into play for relatively painless levels of apparent shock (150volts), implying that the processes underlying obedience for highly immoral actions may be the same as for more mundane forms of obedience.