Question types Flashcards
(22 cards)
Comprehension-Style Questions
The mark scheme for these questions is often very short.
However, to be on the safe-side, don’t write too little and miss out on easy marks.
Think what each of the 3 marks is for? Can you make sure to include all 3 points.
Don’t think that the answer is too obvious, always be as explicit as possible. This is an easy way to pick up 12 marks (if this question style comes up.)
Statistics Questions
Is the finding significant? If so, is it meaningful clinically?
Have the authors used the right statistics?
Are there error bars? What do they signify? Have the authors plotted individual values? If so, what is the spread of the data like?
Have the authors tested for normality? Is data paired or unpaired?
Are the replicates technical or biological?
Is there sufficient power to detect a significant difference?
Usually, if the examiner is asking – there will be something weird about the statistics used. Learn all the appropriate statistics for different scenarios.
Future Experiments
Can you translate the findings to a more relevant model? Longer time-course?
Is there any mechanistic insight offered? If not, can you develop a hypothesis / do hypothesis-generating research?
Is there a need to do more biological replicates?
If there is a knockout, could this be more specific? Could the knockout be done with siRNA or drugs to link to treatment.
Always state the aim and hypotheses of your future experiments, then discuss how the results of your proposed experiments would be useful.
Abstract Writing
Always write the abstract at the end. Use the questions to help you consider the paper in more detail.
Typically 2-3 sentences of an introduction. 1st sentence should outline the broad context, 2nd and 3rd focusing more on the background to the paper.
1 or more sentences on methods, these can be linked with the results. 3-4 results based sentences.
2-3 discussion sentences, placing work in the context. Final sentence should be forward-looking.
Strengths & Weaknesses of a Paper
Don’t be afraid to re-use strengths and weaknesses from earlier questions.
How relevant is the model to the disease being studied?
How well have the authors defined the mechanism behind their findings?
Have the authors shown multiple lines of evidence to support their conclusions?
Are the statistics suitable?
Have all the relevant controls been included? KO confirmed by WB?
Have multiple models been used by the authors?
Critically discuss the data: do what and consider what?
Write what the data shows then consider:
- Have they presented data that supports the conclusions of the Figure?
- Have they included statistical tests and are they appropriate?
- Have they included all appropriate controls (often at fault)?
- Extrapolate to in vivo
- Could they have included anything else to further support their conclusion?
- Are there the correct number of replicates?
- Did they verify a KO/KD?
Identify aim
What were the authors aiming to do?
Usually this is at the end of the introduction.
Write word for word, and mention the method that they used.
Identify aimS
Similar to identify aim.
Give one overall aim (central) then 3 mini aims.
Give some context - why are they aiming for this, what is the importance?
Key figure: do what and consider what?
Select the Figure that best supports the conclusions of the paper and is related to the overall aim. Consider:
- Does it show causality c.f. correlation?
- Is it clinically relevant?
- Which Figure best matches a possible answer to the aims?
- If this Figure wasn’t here, would the paper be any good?
- Will the Figure be useful to a wider audience, or application in other areas?
- Discuss limitations of other Figures
Design an experiment (considerations)
Consider:
- Look at the end to see if they suggested next steps
- Show causation if only correlation has been shown
- Use a better technique
- Provide some insight into the mechanism
- Think clinically
Design an experiment (structure)
State aims and the hypothesis from the start.
Explain the methodology and why you are doing what you are doing.
Suggest the expected results.
Suggest next steps and how it fits into the wider puzzle.
If clinical, outline the treatment regimen.
Comment on
Explain, discuss appropriateness and relate to how it was specifically used in the study.
Assessing statistics (5 considerations)
Have they included error bars?
How many repeats/ data points/ samples per experiment?
Is the data normally distributed?
Have they stated the test, and is it appropriate?
Have they stated what comparisons they are making?
Strengths and weakness (8)
Model (appropriate, multiple in agreement)
Techniques (appropriate, multiple in agreement)
Validity (construct, face, predictive)
Statistical analysis
Ethics
Generalisability possible
Answer question?
Clinical value?
Why may a model be inappropriate?
Lacks the 3 types of validity:
- In vitro
- Genetically modified
- Lab-adapted
- Only shows correlation
- Only shows a modest effect
What can make a model more appropriate?
In vivo
Has all three types of validity
Validated by human data
Uses multiple correlates of pathology
Shows dose dependency
What should be the first things done when looking at a study?
Find what the research question is
Assess the research design
See if the research design appropriately addresses the question
What is a crossover design?
One where subjects receive both intervention and treatment at different times, usually separated by a washout period
What is a parallel group study?
One where 2 groups each receive a different intervention, then results between groups are compared
What is a matched comparison?
One where treatments are compared between those taking it and matched controls
What is a factorial study design?
One where the effects of different variables are studied separately (e.g. in a 2x2 design looking at placebo, each separately, and both together)
