Key Press Issues

Coding key press issues into a .csv spreadsheet

What is a key press issue?

In some of our eyetracking studies, parents wear headphones and repeat sentences to their infant, such as_"Where's the apple?"_ For each trial, when the parent starts saying the target word ("apple"), a researcher presses a key on a controller. The timing of the key press, and therefore the timing of the target word onset, is then registered in the system.

Sometimes, the researcher can't press the key on time or the key press isn't registered. That's why we have key press issues, which we can fix!

How to fix key press issues

[1] Spreadsheet

Open the combined_keypress_issues spreadsheet: Fas-Phyc-PEB-Lab/Seedlings/DSC/DSC_etracking/Key_Press_Issues

The spreadsheet has these columns filled in...

  1. RECORDING_SESSION_LABEL: Subject ID

  2. TRIAL_INDEX: Trial number, including the 4 practice trials

  3. Trial: Trial number, excluding the 4 practice trials

  4. AudioTarget: Target word (e.g., where_apple)

You will eventually fill in these columns...

  1. video_pop_time: Timing of "pop!" sound that prompts parents to say the sentence (in minutes.seconds.milliseconds)

  2. video_targetonset: Timing of target word (in minutes.seconds.milliseconds)

  3. notes

[2] Eyetracking videos

Find the eyetracking videos: Fas-Phyc-PEB-Lab/Seedlings/DSC/DSC_videos

Seedlings videos are here: Fas-Phyc-PEB-Lab/Seedlings/Subject_Files/[subject #]/[month #]/In-Lab_Visit/Eyetracking

Confirm that when you play a video, you can see timestamps in minutes:seconds.milliseconds (00:00.00), not just in minutes:seconds (00:00).

Need help? See the section below titled, "Using the right video program".

[3] Fill in spreadsheet

Open the eyetracking video for a subject with a key press issue.

A test trial begins with a low "bloop" sound, followed by a "pop!" sound that signals the parent to repeat the sentence containing the target word.

In practice trials, the sentences are played over speakers, so the parent does not say anything.

Formatting

In order to get around annoying formatting issues in excel, you should enter times from the video as [minutes].[seconds].[milliseconds]. The video program will likely show this as [minutes]:[seconds].[milliseconds] but excel will round those values in ways we don't want. Examples: ✅ 07.14.03 🚫 07:14.003 You can save over the csv that is currently in the key press issues directory--no need to Save As a new copy.

As you watch the video, note the following...

  1. Trial number. Keep tally so that you know when you're on the trial with the key press issue. In case you lose count, you can write down the target word for each trial.

  2. Once you're on the trial with the key press issue...

    1. Pop time. Pause at the time of the "pop!" You might have to toggle back and forth in the video to get the time correct.

      1. On the spreadsheet, fill in the timestamp for video_pop_time (minutes.seconds.milliseconds, 00.00.00).

    2. Target onset time. Pause once you hear the parent START saying the target word ("a" in "apple").

      1. On the spreadsheet, fill in the timestamp for video_targetonset (minutes.seconds.milliseconds, 00.00.00).

    3. Notes. Fill in if necessary. E.g., "Parent didn't say target."

  3. It can be helpful to follow along with the datasource for the experiment, which should tell you what Trial Index and Trial Number is occurring. These can be located by finding the participant's seedlings control number (search the lab google calendar for their DSC subject id, then look in the calendar event for Seedlings Control number, which will be formatted as subj_month). When you have the control number, go to Fas-Phyc-PEB-Lab/Seedlings/Subject_Files/ and then the subject number and month from the seedlings_control /In-lab_Visit/Eyetracking/DatasourceXXXX.txt. Looking at Trial and AudioTarget should help you follow along in the video (and can help you skip forward if you only have to retrieve a button press from the end of an experiment--just be careful to note that audio targets do repeat, so be *positive* that you're on the correct trial if you do skip ahead in the video).

[4] Q & A

Some common questions, answered.

  • "What if one subject has multiple eyetracking videos?"

    • The video file names indicate their order. Go through the videos in order until you find the trial(s) with key press issues. Usually, when a subject has two videos, the first video records only the practice trials, and the second video records only the test trials.

  • "What if the subject does not have an eyetracking video?"

    • Note this in the notes column on the spreadsheet. Unless a video is found, you do not have to do key press issues for this subject.

  • "What if the parent never says the target word in the trial?"

    • Note this in the notes column on the spreadsheet. Fill in the video_pop_time but leave video_targetonset blank.

  • "What if the parent says the wrong word in the trial?"

    • Note this in the notes column on the spreadsheet. Fill in the video_pop_time but leave video_targetonset blank.

Using the right video program

We need precise timestamps for the timing of the "pop!" sound and target word. Be sure that your computer is capable of displaying video timestamps in minutes:seconds.milliseconds (00:00.00).

Check if these options on your computer display precise timestamps

  1. Your video players.

    1. Right-click on a video, hover over "Open with...", and try the video players listed there.

  2. Your video editors (e.g., iMovie).

  3. Otherwise, install a new video player or editor. See below for suggestions.

Suggestions for video players or editors

Page Status: needs updating

Status details: Duke-related keywords found on the page.

Last updated by ?? on ??/??/??.

Last updated