L5 - L6

Question 1

Surveys

Answer 1

provide stats about a target pop
-process depends on target pop charact
-describe sample -> describe target (meant to be generalized; if not fundamental error)
-responses on opinions, experiences and respondent charact

-good surveys, consistent, reliable and valid
-should be clear
-simple and have no vaugue ?/concepts
-define ambiguous/unfamiliar terms

-comprehend, retrievem judge and respond

-can evaluate via pretesting (which can help for better perfromance…remove ambiious q and improve quality and reduce possible errors in measurement)

-make inferences based off responses

Question 2

Population

Answer 2

all ppl and items with characteristics one wishes to study
-study this pop and make inferences (then use them to generalize back to overall society/pop)

-homo/heterogenity-homo requires smaller samples than hetero (variability)
-nature size-more resources and larger size are better with large pop
-accessibility (can affect choices if inaccessable or hidden)
-spacial distribution (more scattered=more resources needed ie $)

Question 3

pop specification bias

Answer 3

ambiguously derived target pop may lead to this

systematic bias (poor fit below def of target pop/what actually needs to be studied)

Question 4

Sampling

Answer 4

picking a sample from bigger groups (sampling pop) is basis for prediciting prevalence of info/outcome with the group

-simple random (each memb has an = chance of being involved…sampling frame needed; each memb has no and pop units)
-sytematic (every kth unit in a list is selected for inclusion; sampling interval (pop size/desired sample size…ie every 10th…)
-stratified (composing a pop into homogen groups before sampling improves representation (ie age, religion)

-process: prepare for sampling choices, pick census or sampling, choose type (non prob/prob/mixed), determine sample size baed on pop size (large = high confidence…more error margin, lvl, small = low…less error margin) then pick sample

Question 5

Census

Answer 5

data is collected from every memb of pop in/of interest
-not cost effective
-can be dificult

Question 6

Cluster

Answer 6

-multistage
-natural groups/clusters are sampled intially
-members of each selected group are subsampled after (random)
-have desired persons

Question 7

confidence interval

Answer 7

margin of error

level of olerance
-acceptance
0for ammount of error betw sample and pop data

Question 8

hypotheses

Answer 8

re claims in relation to the variables/concepts

-? asked become sources of evidence and invesitgate claims made in hypoth
-need to identify concepts to interrogate hypoth

-get concept, make operational, form indicators and then survey q based on them
-analyze data and determine which is dep and indep and then analize bi or univariately