参数统计和非参数统计是非常重要的对于EEG的分析 , 这篇文章记录了教程中几个tips

1. The ERF data was obtained using ft_timelockanalysis. For the purpose of inspecting your data visually, we also use ft_timelockgrandaverage to calculate the grand average across participants, which can be used for subsequent visualization.

所有数据先被ft_timelockanalysis 处理,然后再将所有trial 利用函数 ft_timelockgrandaverage进行平均,注意教程所提供的数据已经做过ft_timelockanalysis处理。

 

2.FieldTrip does have similar statistical options for frequency data: at the sensor-level we have the ft_freqstatistics function, and on the source-level (statistics on source reconstructed activity), we have the ft_sourcestatistics function, the latter works on data obtained from ft_sourceanalysis).

由于这篇文章基于时域的  但是fieltrip提供在频域的统计分析 比如 ft_freqstatistics 和 ft_sourcestatistics

 

3.In experiments, the data is usually collected in different experimental conditions, and the experimenter wants to know, by means of statistical testing, whether there is a difference in the data observed in these conditions. In statistics, a result (for example, a difference among conditions) is statistically significant if it is unlikely to have occurred by chance according to a predetermined threshold probability, the significance level.

我们的主要目的是要知道是否不同实验组的数据有很大差别,就是说这种差异的随即发生的的概率要小于0.05, 否则很难说是否一定因为数据差异而不是随机或者其他因素导致的

 

4. If the experimenter is interested in a difference in the signal at a certain time-point and sensor, then the more widely used parametric tests are also sufficient.If it is not possible to predict where the differences are, then many statistical comparisons are necessary which lead to the multiple comparisons problem (MCP).A solution of the MCP requires a procedure that controls the FWER at some critical alpha-level (typically, 0.05 or 0.01).

关于使用aprametric或者no parametric ,如果想知道某个时间点 某个Sensor的差异 可以使用参数分析

如果我们不能确定差异来自那个时间点  只能使用 MCP 也就是 no parametric的方式进行分析, 一般我们要求mcp分析的p value小于 0.05

 

5. When parametric statistics are used, one method that addresses this problem is the so-called Bonferroni correction. The idea is if the experimenter is conducting n number of statistical tests then each of the individual tests should be tested under a significance level that is divided by n. The Bonferroni correction was derived from the observation that if n tests are performed with an alpha significance level then the probability that one comes out significantly is =< n*alpha (Boole’s inequality). In order to keep this probability lower, we can use an alpha that is divided by n for each test. However, the correction comes at the cost of increasing the probability of false negatives, i.e. the test does not have enough power to reveal differences among conditions.

如果要使用参数分析  一般是Bonferroni correction, 这种方式很暗揭示内在的区别  不建议使用

 

6.  planar gradient is specific for MEG datasets that have axial/magnetometer sensors (CTF, Elekta)

平面梯度是MEG独有的,不能用在EEG 至于EEG 教程没有单独的代码说明 ,但是我想直接基于timelockanalysis的数据即可

 

7.he format for these variables, are a prime example of how you should organise your data to be suitable for ft_XXXstatistics. Specifically, each variable is a cell array of structures, with each subject’s averaged stored in one cell. To create this data structure two steps are required. First, the single-subject averages were calculated individually for each subject using the function ft_timelockanalysis. Second, using a for-loop we have combined the data from each subject, within each condition, into one variable (allsubj_FIC/allsubj_FC). We suggest that you adopt this procedure as well.

另外关于在使用ft_XXXstatistics函数的时候,数据的保存形式很重要,要符合fieltrip 的输入输出要求,

在本教程中,建议保存cell array structures, 每个Sub的平均数据保存在一个cell, 如何做到?分两步

一,将单个sub的数据进行平均  二,使用for循环讲数据保存成每个特定的实验环境下