update paper.md (citation issue)

paper/paper.md

@@ -40,7 +40,7 @@ The use of wearables in psychophysiology and sports sciences has exponentially i
 
 The oversight regarding the validity of these wearables may be due to the lack of an integrated and user-friendly method for assessing the validity of new wearables. A validation pipeline could lead users from collecting raw data, through pre-processing and advanced statistical analysis, to obtaining the necessary parameters and visualizations required to assess device agreement. The `wearablehrv` package was developed to address this gap. Other existing packages such as `hrv-analysis` [@Champseix2021], NeuroKit2 [@Makowski2021neurokit], pyHRV [@Gomes2019], and similar toolkits offer solutions for pre-processing, analysis, and visualization once IBIs from a single device are provided. However, the added value of `wearablehrv` is rooted in a few key aspects.
 
-First, the currently available packages are not tailored for validation purposes. This becomes especially noticeable when validating multiple wearables at once, across different experimental conditions. This manual process can become a cumbersome task with the packages currently at hand. Second, to make the validation of wearables easier, a user-friendly solution is needed, which current packages lack. This is particularly notable for essential validation steps, such as correcting linear and non-linear lags between devices or trimming signals for a specific device or condition, steps that `wearablehrv` streamlines with GUIs. Third, establishing the validity of a wearable against a gold standard involves numerous decisions. Our pipeline provides a thorough method for users to do this. The availability of such a pipeline also encourages researchers to contribute to establishing a standardized validation protocol, unifying the approach, reducing variability between methods, and facilitating result comparisons. Finally, the division of the `wearablehrv` pipeline into individual and group pipelines offers an advantage for different types of users. The individual pipeline is designed for processing data from a single participant, whereas the group pipeline offers comprehensive tools to establish the quality of the signals, device agreement, and validity of the devices across multiple participants. Most common statistical analyses in validation studies, such as mean absolute percentage error, regression analysis, intraclass correlation coefficient, and Bland-Altman analysis, are already incorporated into the pipeline to streamline the validation process [@BRUTON200094;@d66d4fe8-8a27-315e-bbce-f72f450c6450;@ MAKRIDAKIS1993527;@Haghayegh_2020].
+First, the currently available packages are not tailored for validation purposes. This becomes especially noticeable when validating multiple wearables at once, across different experimental conditions. This manual process can become a cumbersome task with the packages currently at hand. Second, to make the validation of wearables easier, a user-friendly solution is needed, which current packages lack. This is particularly notable for essential validation steps, such as correcting linear and non-linear lags between devices or trimming signals for a specific device or condition, steps that `wearablehrv` streamlines with GUIs. Third, establishing the validity of a wearable against a gold standard involves numerous decisions. Our pipeline provides a thorough method for users to do this. The availability of such a pipeline also encourages researchers to contribute to establishing a standardized validation protocol, unifying the approach, reducing variability between methods, and facilitating result comparisons. Finally, the division of the `wearablehrv` pipeline into individual and group pipelines offers an advantage for different types of users. The individual pipeline is designed for processing data from a single participant, whereas the group pipeline offers comprehensive tools to establish the quality of the signals, device agreement, and validity of the devices across multiple participants. Most common statistical analyses in validation studies, such as mean absolute percentage error, regression analysis, intraclass correlation coefficient, and Bland-Altman analysis, are already incorporated into the pipeline to streamline the validation process [@BRUTON200094;@d66d4fe8-8a27-315e-bbce-f72f450c6450;@MAKRIDAKIS1993527;@Haghayegh_2020].
 
 In summary, provided that a wearable device (either PPG or ECG) allows for the export of the complete time series of recorded IBIs, this Python package makes it relatively easy to establish the validity of a novel wearable in just a few steps. The inclusion of GUI in most functions grants researchers and wearable users the flexibility in validating an unlimited number of wearables across a range of conditions.