My two questions about block average are as below.

(1) The first question is about the baseline of block average. Homer3 says "**The baseline of the average is set to zero by subtracting the mean of the average for t<0**." In our study, the tRange is [-5, 60]. 5 seconds before the test starts and 60 seconds after the test starts to get a good HRF. We repeated three times of the same test. My understanding is "zero" (baseline) is the mean score of HBO during the first 5 seconds before the cognitive test starts. Is this understanding correct?

Next is an example to show my understanding. Let's ignore HBO's unit and suppose the time interval is 1 second.

The first 5 second's original HBO is 8,9,10,11,12, and the mean score of them is 10. Homer does not show the absolute value but the difference between the HBO with the mean score, therefore, we get -2, -1, 0, 1, 2 for the first five seconds. Our data have negative HBO concentration values. However, in the reality, HBO concentration is always positive. I guess the reason is that Homer shows relative HBO values, not absolute values.

Mean score of the first 5 seconds |
Original HBO |
HBO that Homer shows |

10 |
8,9,10,11,12 |
-2, -1, 0, 1, 2 |

(2) The second question is about data analysis. The first 5 seconds' HBO concentration changes are used to set up the baseline. So when we analyze the data, do we need to exclude the first 5 seconds' data, only use data of 60 seconds after the tests start?

Hi Fiona,

Please find my responses below in blue.

- The first question is about the baseline of block average. Homer3 says "
**The baseline of the average is set to zero by subtracting the mean of the average for t<0**." Our tRange is [-5, 60]. 5 seconds before the test starts and 60 seconds after the test starts to get a good HRF. We repeated three times of the same test. My understanding is "zero" (baseline) is the mean score of HBO during the first 5 seconds before the cognitive test starts. Is this understanding correct?

**Homer, as you said, calculates the mean from -5 to 0 (in your case) and subtracts this mean from the whole HRF (i.e. -5 to 60), so that now the new values at baseline i.e. from -5 to 0 are around zero.**

Next is an example to show my understanding. Let's ignore HBO's unit and suppose the time interval is 1 second.

The first 5 second's original HBO is 8,9,10,11,12, and the mean score of them is 10. Homer does not show the absolute value but the difference between the HBO with the mean score, therefore, we get -2, -1, 0, 1, 2 for the first five seconds. Our data have negative HBO concentration values. However, in the reality, HBO concentration is always positive. I guess the reason is that Homer shows relative HBO values, not absolute values.

Mean score of the first 5 seconds |
Original HBO |
HBO that Homer shows |

10 |
8,9,10,11,12 |
-2, -1, 0, 1, 2 |

**In this case yes, the baseline values would be as in your table, the same mean i.e. 10 is also subtracted from the HRF values i.e. 0 to 60. HbO change is not always positive or negative. It depends on your measurements.**

- The second question is about data analysis. The first 5 seconds' HBO concentration changes are used to set up the baseline. So when we analyze the data, do we need to exclude the first 5 seconds' data, only use data of 60 seconds after the tests start?

**If by “using it”, you mean using it for statistical analysis, the answer is yes. You can even crop the first 5 sec after stim to remove the initial rise. E.g. if the duration of your stimulus is 50 sec and you picked the trange as above to also include baseline recovery (-5 to 60). You can take the mean of +5 to 50 and use that value in your statistical analysis.**

Hope this helps!

Meryem