research design Martijn

Question 1

what is a confounding factor?

Answer 1

A variable that affects both the independent and dependent variable

Question 2

3 types of purposes of research

Answer 2

1. exploritive studies
2. descriptive studies
3. explanatory studies

Question 3

Design of simple but effective research:

Answer 3

• improves effectiviness of
  statistics
• avoids wasting time and
  money
• avoids human, animal
  suffering or ecosystem
  disturbance
• minimizes random
  variation
• accounts for confounding
  factors

Question 4

Simpsons paradox:

Answer 4

er is een positieve trend voor elk van twee afzonderlijke groepen, maar er verschijnt een negatieve trend als de data wordt gecombineerd.

Question 5

a hypothesis should always be:

Answer 5

• Be relevant to the research question
• Explain what you expect to happen
• Contain an independent and dependent variable
• Be clear and understandable
• Be testable
• Be measurable

Question 6

control group:

Answer 6

een control group is identiek aan de treatment group. Dit betekent identiek van alle kanten -> ze moten dezelfde strain hebben, zelfde condities en from the same source

Question 7

waarom een control group?

Answer 7

A control offers a baseline against which the
other treatment values can be compared

Question 8

wat is een negatieve controle?

Answer 8

Een negatieve controle is een groep in een experiment die geen enkele behandeling krijgt en daarom geen verandering mag vertonen tijdens het experiment

Question 9

positieve controle

Answer 9

een positieve controle is een groep in een experiment die de benoemde behandeling krijgt, waardoor verwacht wordt dat het een bepaald effect krijgt

Question 10

Concurrent control

Answer 10

A concurrent control is a subject enrolled simultaneously with the treatment group from the same source population and followed for the same study period

Question 11

historical control

Answer 11

a historical control is a subject treated in the past with the standard form of care whose outcomes are used to compare with patients receiving the treatment

Question 12

Pilot studies can:

Answer 12

• Je kennis laten maken met het onderzoekssysteem
• Valideer de biologische vraag (check)
• Oefen en valideer technieken
• Standaardiseer methoden tussen waarnemers

Question 13

Om de hypothese duidelijk te verklaren is nodig:

Answer 13

• strongest evidence
• Controls
• Pilot study & preliminary data

Question 14

difference hypothesis and prediction

Answer 14

A hypothesis explains why
something happens while a prediction suggests that
something will happen

Question 15

random variation (inherent variation, background variation, within-treatment variation, extraneous variation or noise)

Answer 15

any variation in the dependent variable between individuals in our sample that cannot attributed to the independent factors.

Question 16

Replication is important:

Answer 16

-To account for between-individual variation
-To reduce chance when measuring a factor that is high variable (i.e.&raquo_space; SD standard deviation), i.e. extreme outcomes are less likely by chance.
-As random variation gets less (cancelled) when the sample size increases
-As the mean becomes more precise as sample size increases and it becomes easier to discriminate between two populations
-As it allows to estimate variation in populations
-As it allows statistics

Question 17

sample:

Answer 17

to select individuals from the population to be measured

Question 18

Sampling methods:

Answer 18

1)Simple random
2)Stratified
3)Cluster
4)Convenience
5)Self-selection

Question 19

Simple random sampling:

Answer 19

• Use random number
  generator to select
  samples
• Every individual should
  have equal chance of
  being added
• The larger the sample the
  more likely that it will be
  representative
• Sometimes access to data
  of all subjects can be
  difficult or impossible

Question 20

Stratified sampling:

Answer 20

• Depends on variable to
  strongly affect the factor
  of interest
• Requires good
  information on the
  distribution of the strata
  variable

Question 21

Cluster sampling:

Answer 21

• Reduced effort, if
  individuals occur in
  natural clusters
• Be careful not to change
  the sample population
• Pay attention to
  practicality and risk of
  interaction within a
  cluster

Question 22

Problems convenience sampling:

Answer 22

• Selection bias = some
  suitable subjects are
  more likely to end up in
  your sample than others
• Recruitment bias = a
  relation between trait we
  are interested in and
  willingness to be in the
  study
• External validity = the
  extent you can generalize
  your findings to the whole
  population

Question 23

Convenience sampling:

Answer 23

• is not as methodoligical
  as the other sampling
  methods
• its an simple way of
  sampling if you dont have
  enough money or
  recources.
• scottish height of ten year
  olds zoo

Question 24

self-selection

Answer 24

• Limits usefulness (e.g.,
  phone polls are not
  random, and not
  everybody responds to an
  ad)
• In human studies self-
  selection is almost
  inevitable
• Minimize self-selection
  and evaluate the
  consequences

Question 25

pseudoreplicates

Answer 25

non-independent replicates or often called pseudoreplicates

Question 26

What is essential if you want your survey to be independent?

Answer 26

a measurement made on 1 individual should not provide any useful information about that factor on another individual. for example, if you want to know how much parasites blue tits have. This is independent because we chicks in the same box are likely to have similar number of parasites. these are pseudoreplicates