SOC200 - Qualitative Field Research (Chapter 10 + 16) Flashcards
Inferential Statistics & Statistical Significance
whether or not the relationship seen in sample is generalizable to the whole pop
low SE, more confident relationship you are seeing is not due to sampling error
nonprobability methods – high sampling error – misleading statistical significance values, might not accurately reflect relationship
Statistical Significance
refers to chances that relationship observed could actually be a result of a sampling error + therefore doesn’t reflect relationship that actually exists in the population
Statistical Significance
chances high that result of sampling error - not statistically significant
chances low that result of sampling error - statistically significant
Logic behind Statistical Significance
Using EPSEM, accurate conclusions about pop from this sample (no relationship exists)
Logic behind Statistical Significance
not drawn using EPSEM, showing relationship that does not exist in pop
The Dilemma – How do we Decide which Sample Reflects the Population?
Same relationship in the two samples, but only one accurately reflects the true relationship in the population
often don’t know relationship in population
we wouldn’t know which one is accurate
Reduces the issue to two decisions
- sample is unrepresentative (relationship detected in sample appears because sample was not collected representatively (sampling error)
Reduces the issue to two decisions
- sample is representative, reflects an actual relationship in whole population
Statistical Significance Values
various probabilities associated with various degrees of sampling error (sample unrepresentativeness) (# of samples taken, size of sample, assumed distribution of pop parameter)
Statistical Significance Values
Researchers use these probabilities to ascertain whether sampling error in sample is small enough to accept relationship as generalizable to the pop.
statistical significance expressed as a probability: ranges from 0 to 1.
Statistical Significance Values
.05 (p=.05): relationship didn’t really exist in the population, probability of occurring in 5 of 100 samples
lower the p values, the lower the type I error
once sample large enough, everything is gonna be statistically significant
substantive significance: is the differences really meaningful in research?
statistical significance value of .01 (p=.01)/.001 (p=.001)
1 of 100 samples to show relationship seen in sample because of sampling error
1 of 1000 samples to show relationship seen in sample because of sampling error
Field Research
Observing/participating in social behaviour to understand it through deliberate + conscientious collection of data on behaviour
involves the natural activity of observation, most of us have engaged in activities that resemble field research
Field Research
Non-social science researchers (journalists & reporters, social workers, police detectives)
in the social science: greater degree of thoroughness
theory-generating activity – creating a theory based on what you’re seeing
Field Research
anthropologists
can yield quantitative data: # of interactions or behaviours
research right where it happens
deeper understanding of phenomena – greater validity of data
probing life in its natural habitat
observations cover wide range – wherever there’s social interaction
Appropriate Field Research Topics
topics that cannot be easily quantified: change in vocal tone/sudden shift in body language inherent to understanding attitude/behaviour
studying attitudes & behaviours that cannot be understood outside of their natural setting: Native American rain dance
Appropriate Field Research Topics
studying social processes over time: making sense of processes as they happen instead of retrospectively reconstructing them
social development through university experience
several observations of given event/several observations over time
to draw inferences: might require multiple observations
Lofland et al.’s Typology of Appropriate Field Research Topics: Practices
(talking, driving, walking, waiting on someone)
Lofland et al.’s Typology of Appropriate Field Research Topics: Groups & Cliques
small/closely knit groups like sports teams, clubs, work groups, friendship circles
how ppl interact
Lofland et al.’s Typology of Appropriate Field Research Topics: Social & Personal Relationships
studying behaviour “normal” for sets of roles (how are friends supposed to act, how is a father supposed to act towards his son?)
Lofland et al.’s Typology of Appropriate Field Research Topics: Organizations
formal organizations such as work organizations, hospitals, schools, associations, lobby groups, etc.
formal + informal
Lofland et al.’s Typology of Appropriate Field Research Topics: Settlements & Habitats
study of villages, ghettos, neighbourhoods, communities, tribes
Lofland et al.’s Typology of Appropriate Field Research Topics: Subcultures & Lifestyle
studying how large numbers of ppl adjust to life in groups labelled such as “poor”, “rich”, gay, biker, goth, nerd, emo
unique contexts
Lofland et al.’s Typology of Appropriate Field Research Topics: Roles & Social Types
analyzing behaviour associated with positions ppl occupy