RESEARCH SUMMARY SEMESTER 2
All modules
Radboud University, Nijmegen
Made by: Georgia Graat
,Research summary Q3 + Q4
Wet lab research
When doing an intervention study, there are a lot of things you should be able to do on your own. A
study always starts with a good research question. This research question should contain the study
population, study period, study place, other inclusion criteria, the measurement method and the
determinant with outcome.
A really important step is making your own step-by-step protocol of the study. This always contains a
list of the materials you will use. After that, the dilutions for the calibration curve have to be made
and the measurements for this curve. Then you can write down all the steps after that, with
references to protocols and descriptions of things that you did different or wrong.
A calibration curve is a graph with absorption Absorbance creatinine solution at 500
(nm) on the y-axis and the concentration
nm
(mostly mM) on the x-axis. You do
measurements with dilutions with known 1 y = 0,0037x + 0,032
Absorbance at 500 nm
concentrations and measure the absorptions. 0,8 R² = 0,9051
Then you make a formula and the 0,6
corresponding line with it. This line has a R2 that
0,4
shows how much correlation there is/ how far
the line differs from the measured points (R is 0,2
2
Pearson correlation and R is variance). It should 0
be above 0.96 to be a useful calibration curve. 0 50 100 150 200 250
This should be so high, because otherwise the Creatinine concentration (µM)
numbers cannot be trusted. You use it when
you have a solution which you know the absorption of, but not the concentration. You are able to
read and calculate the concentration from the graph. Check if the curve is valid by: checking R2,
taking control samples each time you run the assay and checking slope of line with known values.
Solutions that you use have to be diluted so they fall in the optimal measurement window for the
spectrophotometer and calibration curve. For all dilutions, it is very important to calibrate the
measurement device. For example, the spectrophotometer is calibrated by letting it soak up a blanc
(water) before every measurement. You also need a background or zero measurement that are
‘normal values’ from someone, before doing an intervention, otherwise you can’t spot a difference.
Sometimes, reference values from other studies are also used to compare data. To compare groups,
it is very useful if they have the same starting values.
- Divide standard solution concentration by concentration of wanted dilution. Divide amount
of dilution that you want to make by this number. This is the number of amount of standard
solution you should use and the rest should be water.
- 0.1 ml of a 100x dilution of a compound
o 100 µL compound + 900 µL water (10x diluted)
o 100 µL dilution + 900 µL water (10x diluted the 10x dilution)
▪ Very small numbers can better be diluted in a two-step, because the pipet is
not accurate enough otherwise
- 5 mM from a 40 mM standard solution for 300 µL
o 40/5 = 8 mM and 300/8 = 37.5 µL
o 37.5 µL with 262.5 µL water for 5 mM 300 µL dilution
Made by: Georgia Graat
, There are 3 important methods that are often used to measure for example food intake. In a 24-hour
recall subjects are asked to report all items consumed in the past 24 hours. This is a way to quantify
the intake with a low subject burden, no alteration in normal behavior but with high costs and recall
bias. The diet record is in which subjects record all items consumed over consecutive days where
intake is also quantified. This causes less bias, but has a high subject burden, high costs and may
cause changes in normal behavior. Lastly a food frequency questionnaire (mostly used with large
groups) in which subjects report how frequently certain food items were consumed over a specific
period of time. This has a low subject burden, goes further back in time and doesn’t cause alteration
in normal behavior but has recall bias and is only used for longer periods.
There are a lot of graphs one can use to display results. For instance, the boxplot in which two groups
can be displayed for the same variable. This shows the median, two extreme values and the first and
third quartile. With this, you can easily show the difference within groups. Others are the histogram
and bar chart. A scatter plot is also commonly used to display a correlation between two continuous
variables with the determinant on the x-axis and the outcome on the y-axis. When drawing this, you
should draw the line of correlation and all dots that represent the individual results. Draw a scale.
In a scatter plot, one can measure the Pearson correlation. This is the strength of the correlation
between two continuous variables. -1 and 1 are perfect correlations, whilst 0 is absolutely nothing.
You can also look at the significance 2-tailed. This is the p-value you set for yourself in a study. Mostly
used is 0.05, which means there is a 5% chance for error. If the significance is below the p-value, you
can assume the association that you found is significant. This says sort of the same as the 95%
confidence interval for a difference that should not contain 0. When this happens, the results are
also significant, because the difference is never 0 meaning the 2 groups clearly differ in means.
Bias and validity are each other’s opposites and display systematic errors in the study. This means
there is an error in the measurement or information that causes the results to always lean towards a
higher or lower number than in reality. For example, a measuring tape that starts at 3 cm always
measures 3 cm too much. Results are more precise (and reliable) when all measurements are around
the same number, so little spread, and results are valid when there are no systematic errors. To
cover the precision, measurements are usually done in triplo (measuring everything 3 times). These
triplo measurements should be relatively close together, otherwise you introduce large
measurements errors.
Information in a report should be carefully written down and examined and all should be in a specific
order. The title should contain information about the conclusion, study population and study time.
The introduction covers all general background that we know about the determinant, measurement
method and outcome in the particular study. It should also lead to a research question. You should
also explain the chosen approach; why these measurement methods and materials. In the materials
and methods section, you tell about the study population, measurement materials, databases,
programs used, protocol with a small summary of what was done. The results are only what you have
observed and what is shown in the tables and figures. The figures should include axes with correct
units, a title, a legend and correct results. The discussion tells about the conclusions that can be
drawn from the results and if this is correlated to the background information. It can also criticize the
measurement methods and protocols used and give implications for further research. After that,
there is one short answer to the research question with maybe an implication on the significance,
clinical impact and reliability of the study. The references are always in Vancouver.
A biomarker is a measurable indicator of some biological state or condition. It is suitable biological
material (urine, blood or tissue) which quantity is related to the amount of molecule of interest. A
Made by: Georgia Graat
All modules
Radboud University, Nijmegen
Made by: Georgia Graat
,Research summary Q3 + Q4
Wet lab research
When doing an intervention study, there are a lot of things you should be able to do on your own. A
study always starts with a good research question. This research question should contain the study
population, study period, study place, other inclusion criteria, the measurement method and the
determinant with outcome.
A really important step is making your own step-by-step protocol of the study. This always contains a
list of the materials you will use. After that, the dilutions for the calibration curve have to be made
and the measurements for this curve. Then you can write down all the steps after that, with
references to protocols and descriptions of things that you did different or wrong.
A calibration curve is a graph with absorption Absorbance creatinine solution at 500
(nm) on the y-axis and the concentration
nm
(mostly mM) on the x-axis. You do
measurements with dilutions with known 1 y = 0,0037x + 0,032
Absorbance at 500 nm
concentrations and measure the absorptions. 0,8 R² = 0,9051
Then you make a formula and the 0,6
corresponding line with it. This line has a R2 that
0,4
shows how much correlation there is/ how far
the line differs from the measured points (R is 0,2
2
Pearson correlation and R is variance). It should 0
be above 0.96 to be a useful calibration curve. 0 50 100 150 200 250
This should be so high, because otherwise the Creatinine concentration (µM)
numbers cannot be trusted. You use it when
you have a solution which you know the absorption of, but not the concentration. You are able to
read and calculate the concentration from the graph. Check if the curve is valid by: checking R2,
taking control samples each time you run the assay and checking slope of line with known values.
Solutions that you use have to be diluted so they fall in the optimal measurement window for the
spectrophotometer and calibration curve. For all dilutions, it is very important to calibrate the
measurement device. For example, the spectrophotometer is calibrated by letting it soak up a blanc
(water) before every measurement. You also need a background or zero measurement that are
‘normal values’ from someone, before doing an intervention, otherwise you can’t spot a difference.
Sometimes, reference values from other studies are also used to compare data. To compare groups,
it is very useful if they have the same starting values.
- Divide standard solution concentration by concentration of wanted dilution. Divide amount
of dilution that you want to make by this number. This is the number of amount of standard
solution you should use and the rest should be water.
- 0.1 ml of a 100x dilution of a compound
o 100 µL compound + 900 µL water (10x diluted)
o 100 µL dilution + 900 µL water (10x diluted the 10x dilution)
▪ Very small numbers can better be diluted in a two-step, because the pipet is
not accurate enough otherwise
- 5 mM from a 40 mM standard solution for 300 µL
o 40/5 = 8 mM and 300/8 = 37.5 µL
o 37.5 µL with 262.5 µL water for 5 mM 300 µL dilution
Made by: Georgia Graat
, There are 3 important methods that are often used to measure for example food intake. In a 24-hour
recall subjects are asked to report all items consumed in the past 24 hours. This is a way to quantify
the intake with a low subject burden, no alteration in normal behavior but with high costs and recall
bias. The diet record is in which subjects record all items consumed over consecutive days where
intake is also quantified. This causes less bias, but has a high subject burden, high costs and may
cause changes in normal behavior. Lastly a food frequency questionnaire (mostly used with large
groups) in which subjects report how frequently certain food items were consumed over a specific
period of time. This has a low subject burden, goes further back in time and doesn’t cause alteration
in normal behavior but has recall bias and is only used for longer periods.
There are a lot of graphs one can use to display results. For instance, the boxplot in which two groups
can be displayed for the same variable. This shows the median, two extreme values and the first and
third quartile. With this, you can easily show the difference within groups. Others are the histogram
and bar chart. A scatter plot is also commonly used to display a correlation between two continuous
variables with the determinant on the x-axis and the outcome on the y-axis. When drawing this, you
should draw the line of correlation and all dots that represent the individual results. Draw a scale.
In a scatter plot, one can measure the Pearson correlation. This is the strength of the correlation
between two continuous variables. -1 and 1 are perfect correlations, whilst 0 is absolutely nothing.
You can also look at the significance 2-tailed. This is the p-value you set for yourself in a study. Mostly
used is 0.05, which means there is a 5% chance for error. If the significance is below the p-value, you
can assume the association that you found is significant. This says sort of the same as the 95%
confidence interval for a difference that should not contain 0. When this happens, the results are
also significant, because the difference is never 0 meaning the 2 groups clearly differ in means.
Bias and validity are each other’s opposites and display systematic errors in the study. This means
there is an error in the measurement or information that causes the results to always lean towards a
higher or lower number than in reality. For example, a measuring tape that starts at 3 cm always
measures 3 cm too much. Results are more precise (and reliable) when all measurements are around
the same number, so little spread, and results are valid when there are no systematic errors. To
cover the precision, measurements are usually done in triplo (measuring everything 3 times). These
triplo measurements should be relatively close together, otherwise you introduce large
measurements errors.
Information in a report should be carefully written down and examined and all should be in a specific
order. The title should contain information about the conclusion, study population and study time.
The introduction covers all general background that we know about the determinant, measurement
method and outcome in the particular study. It should also lead to a research question. You should
also explain the chosen approach; why these measurement methods and materials. In the materials
and methods section, you tell about the study population, measurement materials, databases,
programs used, protocol with a small summary of what was done. The results are only what you have
observed and what is shown in the tables and figures. The figures should include axes with correct
units, a title, a legend and correct results. The discussion tells about the conclusions that can be
drawn from the results and if this is correlated to the background information. It can also criticize the
measurement methods and protocols used and give implications for further research. After that,
there is one short answer to the research question with maybe an implication on the significance,
clinical impact and reliability of the study. The references are always in Vancouver.
A biomarker is a measurable indicator of some biological state or condition. It is suitable biological
material (urine, blood or tissue) which quantity is related to the amount of molecule of interest. A
Made by: Georgia Graat
