Literature Review and Study Design

Question 1

What is a research question?

Answer 1

A research question provides focus for an investigation
-defines an area of interest
fundamental core of any research project

Question 2

What are the five key points for a good research question?

Answer 2

1. Feasible
2. Interesting
3. Novel
4. Ethical
5. Relevant

Question 3

What is a hypothesis?

Answer 3

A hypothesis should describe what you expect to happen in your study
provides a reference point for study design
reference point for discussion

Question 4

What is a null hypothesis?

Answer 4

states nothing will happen, that there will NOT be an effect.
* e.g.: There is no difference in owner reported aggression between castrated and uncastrated dogs.

Question 5

What is an alternate hypothesis?

Answer 5

states what you expect to happen, that there WILL be an effect.
* e.g.: There is an increased incidence of owner reported aggression in uncastrated dogs compared to castrated dogs.

Question 6

What is a two-sided hypothesis?

Answer 6

There is a difference between group means (either increase or decrease)

Question 7

What is a one-sided hypothesis?

Answer 7

There is a difference between group means in a given direction.
* Only used if there is some theoretical reason why any differences could be in one direction.

Question 8

What is the population?

Answer 8

Complete set of people with specified characteristics

Question 9

What is the ‘target population’ ?

Answer 9

Clinical and demographic characteristics define the target population

Question 10

What defines the accessible population?

Answer 10

Geographical and temporal characteristics define the accessible population

Question 11

What is the intended study sample?

Answer 11

is the subset of the
accessible population that is sought to be included
in the study

Question 12

What is the actual study sample?

Answer 12

is the group of subjects
that do participate in the study

Question 13

What is random sampling?

Answer 13

The population & sample size is defined
» Population must be larger than the sample size required

Question 14

What is the benefit of random sampling?

Answer 14

Good validity of generalising the findings from a sample to a population

Question 15

What is systematic sampling?

Answer 15

Resembles a simple random sample by enumerating the population but then the
sample is selected by a preordained periodic process

Question 16

What is stratified sampling?

Answer 16

Divide the population into subgroups according to characteristics
* Sex
* Age
* Severity of clinical characteristics
» Take a random sample from each of these strata

Question 17

What is proprtionate stratified sampling?

Answer 17

Number of subjects selected depends on size of strat

Question 18

What is disproprtionate stratified sampling?

Answer 18

Equal number of subjects from each strata irrespective of size of
each strata.

Question 19

What is cluster sampling?

Answer 19

A random sample of natural groupings (clusters) of individuals within the population
» Useful when the population is large and widely dispersed

Question 20

What are the multiple stages of cluster sampling?

Answer 20

Random sample of pig farms
* Random sample of pens within each pig farm
* Radom sample of individuals from each pig pen selected

Question 21

What is the definition of inclusion criteria?

Answer 21

These define the main characteristics
of the target population that pertain to
the research question.
* All subjects must fulfill these
requirements to enter the study.

Question 22

What is the definition of exclusion criteria?

Answer 22

reasons why a
subject should not be enrolled in a study.
» For example:
* High likelihood of being lost to follow-up
* High risk of side effects
* Characteristics that make it unethical to
withhold additional treatment

Question 23

What are the two main goals of recruting subjects into a study?

Answer 23

Recruit a sample that represents the target population
* Minimise the prospect of getting the wrong answer due to systematic error (bias).
» Recruit a sufficient sample size
* Minimise the chance of getting the wrong answer due to random error (chance).