No announcement yet.

Reducing "bad" posturography data

  • Filter
  • Time
  • Show
Clear All
new posts

  • Reducing "bad" posturography data

    I recently collected static balance data on a group of dancers in various positions (dance specific poses like arabesque, releve, front attitude) and conditions (eyes open/closed, foam/firm surfaces). In the eyes open stable surface conditions participants typically did a great job of maintaining their balance and not needed major postural corrections (like a step down) during trials. However, in the more challenging conditions, participants often touch down to regain their balance. In analyzing the center of pressure data, I'm not sure how to handle trials in which participants lost their balance and put a foot down. We are planning to look at a mix of global and structural metrics of CP data (standard deviation, mean velocity, ellipse area, sample entropy, DFA alpha) and I'm concerned filtering out the touchdowns is not the best method to deal with this. Additionally, only looking at a small sample of the trial that has no touchdowns is not an accurate picture of the participant's behavior. I'd like to use the entire 20-second trial if possible, but open to ideas!

    Any guidance and suggestions will be greatly appreciated!

    Thank you,
    Matthew Wittstein
    Elon University