How Science Works Flashcards
What does standard deviation show?
Spread of data around a mean and effect of any outliers
Why should you use standard deviation bars are not ranges
- Allows us to see the spread of data around a mean
- Allows us to see effect of any outlines.
What can scientist do to make this more reliable?
- Repeat
- Stats test to see if significant differences are due to chance
- Peer review
Valid conclusions need to beβ¦.
- Large sample size so representative
- Control used to see the effect ofβ¦
- Mean and SD to determine significant differences
Null hypothesis
X will have no effect if on the rate of reaction of y therefore
There will be no significant between mean rates of reaction
Explain why a log scale is used
Large range in numbers and allows us to see exponential growth
Large SD bars shows
A large range in variation around the means so less reliable mean value
Small SD bars show
A small range of variation around the means so more accurate mean value
Suggest why scientists recorded both percentage change and tumour volume
Percentage change = slows us to compare tumour as may differ in size/ volume
Tumour volume = as tumours may differ in shape
AO3 PLAN
- 2-3 comments on general trend
- Does STUDENTS CONCLUSION match data
- Validity = Where was the experiment data from ( computers,cultured,in test tube, not in sane environment so act with different behaviour, Petri dish ( in vitro ), ISOLATED
- If medicine, do we know dosage, side effects
- No stats test to see significant differences
- Other abiotic / biotic factors that effect data
- Reliability = sample size, length of study, season , habitat
- Is there a Control to compare to
When do u use t test
To compare 2 means to see if they are significantly different
When do u use correlation coefficient
To see if link between 2 covariables is significant
When do you use chi squared test
To see if differences in observed and expected values in categorical data are significantly different
P< 0.05
- There is a significant difference between mean rate of reaction
- There is a significant increase / decrease in rate of reaction from
- There is a less than 5% probability that significant differences are due to chance
- Reject null hypothesis
P>0.05
- There is no significant difference in mean rate of reaction.
- There is more than 5% probability significant difference are due to chance
- Accept null hypothesis
P<0.01
- There is a highly significant difference between mean rate of reaction
- There is a significant increase/ decrease in mean rate of reaction at β¦compared toβ¦
- There is less than 1% probability that significant differences are due to chance
- Reject null hypothesis
R= 1, P<0.05
- There is a strong positive correlation betweenβ¦.
- As β¦ increases , the other factor increases
- This relationship is significant as P<0.05, meaning less than 5 % probability that the relationship is due to chance
- Reject null hypothesis
R=-1, P>0.05
- There is a strong negative correlation between x And y
- as x increase , y decreases
- The relationship is not statistically significant as P > 0.05 which means greater than 5 % probability that the relationship is due to chance
- Accept null hypothesis
SD bars overlap
There is a significant difference
( direction ) significant decrease in mean rate of reaction from 20 to 40
SD bars donβt overlap
There is a significant difference
( direction ) significant increase in mean rate of reaction from 20 to 40
What is an control used for
To allow comparison to the the effect of
Describe how to set up a control
Set up a control with no ( to the effect of ) independent variable with all environmental factors the same ( temperature, light, PH)
Degrees of freedom
N-1 ( number of different types of β¦ -1)
How to determine significant difference from chi squared table
Go down column with P=0.05
Go across column with degree of freedom value
If x2 value of greater than expected value than no significant difference, not due to chance