Task 3 Making Observations and asking questions Flashcards
Reliability
Reliability of a measure concern its ability to produce similar results when repeated measurements are made under identical conditions
Validity
Validity of a measure id the extent to which it measures what you intend it to measure
Face validity
descries how well a measurement instrument appears to measure what it was designed to measure
Content validity
has to do with how adequately the content of a questionnaire sample the knowledge, skills, behaviour that the set is intended to measure (covering the knowledge which you have done the questionnaire for)
criterion-related validity
reflects how adequate a test score can be use to infer an individuals value on some criterion measure (how well does the score predicts future outcomes of a test about the same topic)
concurrent validity
if the scores and the criterion are collected at the same time
predictive validity
comparing the scores of your test with the value of criterion measure observed at an later time
Construct validity
Validity of a test judged by how adequate a test samples behaviour representative for the universe of behaviours the test was designed to sample
Test-retest reliability
test two times with a long time in between
→when correlations are high your measure is reliable
Parallel-forms reliability
same as test-retest but the second time you will get a parallel form which fights the problem of recalling
Split-half reliability
parallel test in the same session
Interrater reliability
A level in which multiple observers agree in classification and quantification of behaviour
preliminary observations
observe your objects in the certain situation before the real study and list all their behaviour
literature search
means to look at previous experiments in your field and try to copy the behavioural categories you should do preliminary observations as well to be sure that they fit
Quantifying behaviour
- Frequency method
→You record the number of times a behaviour occurs in a certain time - Duration Method
→how long does a particular behaviour last - Intervals method
→create time intervals and look if the behaviour occurs within the interval (make them so short that only one behaviour can occur per interval)
Time sampling
scan the group for a certain time an than record the observed behaviours for the same time (making more detailed records of your observation)
Individual sampling
You choose a single subject and observe it for a period and record the observations than you choose another subject and repeat the procedure (when you want to preserve the organization of an individual’s behaviour)
Event sampling
You observe only one behaviour (e.g. sharing behaviour) and record all instances of that behaviour (when you clearly define one behaviour of interest)
Naturalistic observations
observing your objects in their natural environment without manipulating variables
Unobtrusive observations
observations that do not alter the natural behaviours of your subjects (stay hidden or let the objects get used to you)
Ethnography
a researcher becomes immersed in the behavioural or social system being studied
→gaining access to a field setting
→gaining entry into the group
→becoming and staying invisible
Making observations and recording data
Easy if you are not covert but when you are take short notes
Analysing ethnographic data
→1reading your field notes and identify any themes or hypothesis,
→2code any systematic patterns in your notes and consider doing an in depth content analysis
sociometry
involves identifying and measuring interpersonal relationships within a group(friendship patterns among children, peer assessments of teenagers solicited to dealing drugs, as well as other social network and work relationships) the participants have to fill out a questionnaire as well
Case history
is a descriptive technique in which you observe and report on a single or few case/s. (case is the object of study, such as the development of a mental disease of a certain individual)
special kind of demonstration
Archival research
studying existing records (police crime reports, published research articles, social media information)
→should have a specific research question
Content analysis
• To analyse a written or spoken record for the occurrence of specific categories or events (pauses in a speech), items(negative comments, or behaviour)
Performing content analysis
→define response categories
→read or listen to your material first
→record unit =frequency of a word or sentence
The reactive nature of psychological measurement
subjects may behave different when they know they participate in a study
→Human: will interpret your study and behave like they think it might be right
Expectancy effects
when a researcher preconceived ideas about how participants should behave are subtly communicated to subjects and, in turn, affect the participants behaviour
Single-blind technique
the experimenter does not know which treatment the participant take part in so he cant assume behaviour
Double-blind technique
neither the experimenter nor the participant knows with treatments the participant is receiving (e.g. placebo)
Automating your experiments
eliminate some source of experimenter bias, can safe time, tend to be more accurate because it doesn’t make faults humans do
conducting a pilot study
small scale version of the study used to establish procedure, material, and parameters of the real study, needs a lot of time and can be expensive
Manipulation check
You want to check if the participant understood the manipulation right or if he/she reacted different to the manipulation
Field survey (good questionaire)
directly question your participants about their behavior and their underlying attitudes, beliefs, intentions > remain anonymity and confidentiality of participants
Demographics
characteristics of participants (e.g. sex, age,..)
Predictor variable
Demographics are used as predictor values when they determine whether particular characteristics correlate with or predict responses to other items in the survey
Criterion variable
items that are designed to assess the behavior of interest
Open ended items
allow participants to respond in their own words
Restricted Items
provide a limited number of specific response alternatives
Partiallyopenended Items
resemble restricted items but provide an additional other space to give an answer not listed among the specific alternatives
Rating Scales
provides a graded response of the question
Likertscale
provides a series of statements to which participants can indicate degrees of agreement or disagreement
Non response bias
When people don’t answers a survey
4 contact e-mail
Informing e-mail some days in advance
Cover letter contains request to do the survey/contains possibility for questions/signature of you
Thank you card: Thanks for doing it or reminding to do it
Reminding letter reminds people to fill out the questionnaire (only sent if it not already done)
Role attitude cues
are unintended cues in an experiment that suggest the participant how to they are expected to behave
→cooperative attitude: desire to please the experimenter (can cause bizarre behaviour)
→apprehensive attitude: when participants are worried about what will happen to them
→negative attitude: try to ruin the experiment (can be caused by required participation)
Behavioural measures
you record actual behaviour of your subjects
→Frequency: you count how often a behaviour occurs
→Latency: You measure the amount of time a subjects need to respond to some stimulus
→Number of errors: Counting the number of errors
Physiological measures
→requires special equipment to monitor the participants bodily functions
→non-invasive and do not harm the participant
→provides you with accurate information about such things as the state of arousal within the participants body
Q-sort methodology
involves establishing evaluative categories and sorting items into those categories
→retrospective verbal report: participants evaluate on past experiences
→prospective verbal report: participants evaluate how they would behave in a certain future situation
Demand characteristics
cues inadvertently provided by the researcher, participants start to create their own hypothesis and act in certain ways
Facilitated communication
a way to allow severely impaired persons to communicate with others in ways and at levels previously believed to be impossible (body behaviour and gestikulierung)
Probability sampling
Everybody has the same probability to get chosen
Representative sample
should display the average thoughts of the group
Simple random sampling
random number how your subject is chosen
Segment sampling
choose an equal amount of every segment
Cluster sampling
use natural groups in a bigger group (subgroups)