weeks 1,2 and 3

Question 1

where do research questions come from?

Answer 1

• observations
• personal interest and curiosity
• scientific literature
• practical need for more information
• answers to earlier research questions

Question 2

scientific method

Answer 2

1. identify an interesting topic or observation
2. define a question
3. define aims and objectives for research
4. formulate hypothesis
5. collect data to test hypothesis (reproducibly)
6. data analysis to test hypothesis (reproducibly)
7. interpret results and draw conclusions
8. publication and re-testing

Question 3

what makes research questions good?

Answer 3

• relevant to topic of interest.
• address a knowledge gap.
• compelling - why is it wanted?
• answerable through data collection and analysis
• focused - guides efforts of the researcher and sets clear parameters.

Question 4

what is the aim of a study?

Answer 4

• states why the research is being done
• describes a specific end that the study works towards
• leads to achievement of goals

Question 5

what is the objective of a study?

Answer 5

• states how the aims will be achieved.
• should be specific and measurable
• each objective should be achieved using quantitative evidence
• objectives affect the analysis we undertake and thus the data we should gather.

Question 6

hypotheses

Answer 6

• testable statements about the system.
• gives rise to a prediction which could be proven false through collections and analysis of data.

Question 7

four categories of quantitative goals of studies

Answer 7

1. assessment/quantification - assessing current status of a system
2. relationship between variables - variation in data is the basis for inferring relationships between variables.
3. causality - inferred through design, analysis and interpretation of controlled experiments
4. system prediction - depends on accurate and precise info about each system component.

Question 8

what are some practical constraints?

Answer 8

1. feasibility - can it currently be done, technically?
2. time - is it urgent? deadline?
3. knowledge - experience and expertise
4. availability - site accessibility.
5. resources - financial, human, physical.
6. ethical issues.

Question 9

how does bias affect data?

Answer 9

precise but not accurate

Question 10

how does noise affect data?

Answer 10

accurate but not precise.

Question 11

what happens to data when there is bias and noise at the same time?

Answer 11

not accurate and not precise

Question 12

why do we sample?

Answer 12

provides an estimate for total population to avoid a potentially impractical and destructive census.

Question 13

factors to take into account when sampling.

Answer 13

• every individual in a population is different from all others.
• every location is different from all others
• every sample mean will be different from every other.

Question 14

why is the mea of samples of interest?

Answer 14

provides an estimate of population mean.

Question 15

factors affecting precision

Answer 15

1. machine/equipment or observer
2. sampl to sample (sampling error)
3. between different environments and habitats
4. between sampling units

Question 16

how to manage spatial variation?

Answer 16

• ignore (affecting precision)
• impose boundaries (affects generality)
• measure - i.e. treat as covariate
• use as basis for stratification (find a way of categorising the study area according to its key features)

Question 17

sampling options

Answer 17

random
stratified

Question 18

simple random sampling

Answer 18

• sampling at randomly selected coordinates.
• reduces effort as only sampling smaller proportion of entire population.
• inefficient as difficult to figure out where to sample and where not to sample which could take more time.

Question 19

systematic smapling

Answer 19

sampling at regular intervals across entire population.

representative and efficient

Question 20

how can systematic sampling be risky?

Answer 20

may be some underlying factor which has the same periodicity as the sampling structure.

Question 21

stratified random sampling

Answer 21

representative and efficient.

sampling targeted towards finding estimates for each part of site and then combining these to gather overall estimate.

estimate gathered for each group.

Question 22

smapling bias

Answer 22

the difference between the population mean and the sample mean - an unrepresentative sample.

Question 23

sources of bias

Answer 23

• machine/equipment issues
• observer bias

Question 24

what are the multiple sources of uncertainty that could arise when estimating total biomass of forests?

Answer 24

1. spatial variation of trees within sample plot.
2. spatial variation among multiple sample plots
3. noise around data (residual)
4. particular sample of tree population used - coefficients.
5. measurement of stem diameter

Question 25

how can you manage multiple sources of error?

Answer 25

calculating with a range of values instead of a mean gives you a range of results. - understand the formula involved and create a range of possible results for each point of uncertainty. - the error propagates.