Lecture2 - sampling Flashcards

1
Q

describe the differences between a population and an individual

A

An individual animal either has the disease or not, whereas in a population animals with or without the disease may be present
also
an individual animal shows certain signs of the disease, whereas in a population these signs may differ between animals

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Prevalence and distribution are characteristics of …?

These characteristics cannot be established by examining …?

A

the population

an individual animal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

define sampling

A

method to collect information about a subset of the population to make inferences about the characteristics of that population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

why would you choose to sample over running a census?

A

census can be: costly, take forever, unfeasible, may contain irrelevant differences
a sample can: generate results faster, data less expensive to collect, sample results may be more accurate as it is possible to make more efficient use of resources

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

define target population

A

population to which the study results are to be extrapolated

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

define source population

A

the immediate population for which the study conclusions are to be used and from which the subset is to be selected

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

define sampling frame

A

lists all sampling units in the source population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

define sampling units

A

individual members of the sampling frame

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

define internal validity

A

describes the relationship between study and source populations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

define external validity

A

describes the relationship between the source and target populations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

draw flow chart with terminology used in sampling process

A

fig 7.1 text book or lecture slide 22

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Sampling design has 2 factors, what are they and what do they imply?

A

methodology (bias)

sample sizes - statistical inference (precision)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What are the general considerations for a study design? use injuries of working dogs in Aus as eg

A
  • state objectives clearly and concisely
    eg. estimation of the prevalence of injuries in working dogs in Australia
  • Define the target population
    eg. All working dogs in Australia
  • Define the study population
    eg. All working dogs on farms in the Riverina
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

what are the main types of sampling?

A

Non-probability sampling

Probability sampling

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

define non-probability sampling

A

not based on random sampling techniques

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

define probability sampling

A

based on random sampling techniques

17
Q

name the methods within non-probability sampling

A
  • convenience
  • judgement
  • purposive
18
Q

name the methods within probability sampling

A
  • simple random sample
  • systematic random sample
  • stratified random sampling
  • cluster sampling
  • multi-stage sampling
19
Q

define convenience sampling. give eg.

A

sample is selected because it is easy to obtain

eg. estimate prevalence of diarrhea in calves
- > select cows from herds owned by nearby farmers

20
Q

define judgement sampling give eg.

A

“representative” units of the population are selected
eg. estimate prevalence of diarrhea in calves
-> select calves that the investigator thinks are
representative (born within last 15 days)

21
Q

define purposive sampling. give eg.

A

selection is
based on known exposure or disease status (ie. targets specific risk groups)
eg. association between diarrhea in calves and farms which buy replacement cows (cohort study) -> Select herds on known exposure

22
Q

what is the main disadvantage of non-probability sampling?

A

“representativeness” can not be quantified

23
Q

what are the principles of using probability (random) sampling techniques?

A
  • sample representative of target population (small to no bias)
  • estimate of precision possible: standard error of the mean
  • Assumption that any of the possible samples from the source population has the same chance of being selected
24
Q

Describe how simple random sampling (SRS) works

pros v cons?

A
  1. Assign each sampling unit a number 1 to N
  2. Pick a sample of n of these units using a formal random process (random tables)
  3. Sample size calculations based on assumption of SRS
    pros: - Every unit has the same probability of being selected
    - simple concept
    cons: - Not commonly used, it’s not practical
    - Sampling frame often not available with animal populations
25
Q

how would you sample 100 herds from 1000 herds using simple random sampling (SRS)?

A

number herds between 1 and 1000, then choose 100 randomly chosen (by computer or ‘eyes closed’) herds would represent a simple random sample

26
Q

Describe how systematic sampling (SS) works

pros v cons?

A

All sample units (e.g. animals) should be subject to a sequence (e.g. lined up!)
study units are selected at fixed intervals where:
sampling interval = study group size/sample size (k=N/n)
pros: - easy to apply in field conditions
cons: - May introduce bias if characteristic measured is related to sampling interval
- DO NOT USE Systematic Sampling if there is periodicity or ordering in the sampling frame

27
Q

how would you sample 15% of 100 cows using systematic sampling (SS)?

A

k=N/n
N = 100, n = 20 then k would = 100/20 => 5
choose a number

28
Q

Describe how stratified random sampling works

pros v cons?

A

factors likely to influence outcome are identified and used PRIOR to selecting the sample, to divide the sampling frame into strata. Then a simple random sample (SRS) or systematic sample (SS) is selected within each stratum.

pros: - improves precision of the estimate
- assures that all strata are included in the sample
- more flexible - different sampling % in various strata
cons: - must know beforehand to which strata each sampling unit belongs

29
Q

How would you use stratified random sampling in a herd of 7800 chooks if sample size is 60 but they’re in 4 different pens containing different numbers per pen? eg. pen 1 = 3000chooks, p2 = 800, p3 = 2500 & p4 = 1500

A

60/7800 x 3000 = 23 sample size for 1st strata
60/7800 x 800 = 6 sample size for 2nd strata
60/7800 x 2500 = 19 sample size for 3rd strata
60/7800 x 1500 = 12
total = 60

30
Q

Describe how cluster sampling works

pros v cons?

A

involves sampling groups or clusters of individuals (litter, pen, herd) all individuals within cluster are tested
selection of clusters by:
- simple random sampling
- systematic sampling
- stratified sampling
pros: - easy to select
- practical for field use (thus often used)
cons: - information is biased!
- estimate is less precise ie. The occurrence of disease, especially of infectious disease, varies more between clusters than within clusters