Session 5 - PsychoPy Flashcards
Neuroscientists often want to present stimuli to subjects. These can be
pictures, words, sounds or more complex things like pulses of pain, electric shocks, smells or vibrations.
There are many computer programmes that can help you make and present stimuli on a computer screen such as
Psychopy
There are many computer programmes that can help you make and present stimuli on a computer screen. Here are some that we have used in the past:
ePrime
Presentation
Psychtoolbox
Psykinematix
Unity
They all have strengths and weaknesses. Many of them are
expensive, some require significant understanding of a particular programming language (Psychtoolbox = MATLAB, Unity = C#)
Psychopy is piece of software developed in Python by
scientists at Nottingham.
Why we teach Psychopy? - (3)
1: It is free. This means that lots of labs can use it and that you can personally have a copy on your own machine at no cost. In fact we encourage you to download it here: https://www.psychopy.org/download.html. You can get all the older versions (useful for compatibility) here: https://github.com/psychopy/psychopy/releases
2: It uses Python. And Python is the language that we are using for this course.
3: It is ‘careful’. Psychopy is written by vision scientists (among others) and these people need to know lots of details about the screen and the computer. They care about tiny timing errors of tens of milliseconds and how big something looks on the screen. Some of the other packages are a little more ‘loosey-goosey’.
Diagram of Builder view of Psychopy:
In Builder view, This is a way to build experiments
without coding
Why is it good to use psychopy to build experiments without using code? - (2)
or one thing, you need to know what experiments look like before writing one yourself - a ‘high level’ view is important.
But also, the Psychopy builder is a great way to make stimuli! The people who make Psychopy ensure that it runs more or less without bugs and that it does all the right things. And you can inspect and edit the code that it generates later anyway.
Diagram of Psychopy builder view labelled
What does ‘components’ section do?
The things we can add to a routine e.g., a text, picture, key or mouse responses
What does ‘flow’ section do? - (2)
Shows timeline of the experiment
Shows order in which things occur and if they repeat in a loop
What does ‘routine’ seciton do? - (2)
What happens in a single routine e.g., show some text, wait for a key press
Several routines can be added to the flow
What is a routine?
basic building block of Psychopy experiments
What does this blue box represent in flow?
A trial ‘routine’
What will this experiment do?
Show a blank screen for a certain duration
There are many different components in Psychopy such as - (8)
‘T’ = present text stimuli like for instructions and ‘get ready’
Sound: Play recorded files
Keyboard: Respond with keypress
Mouse: Respond with mouse press
Images: Images to display during experiment
Slider: Use sliders during the experiment to get participants rating
Can also get a range of stimuli like gratings, movies etc also range of response devices
In custom zone there is also ‘code’ which allows you to slot python code in experiment to control experiment and produce variables
What does clicking text from components show?
The ‘text’ has certain properties
You can edit the properities of the ‘text’ stimuli by
double clicking
What do these text properities show? - (3)
Name: name the text stimuli so far its named as ‘text’ which will show inside the routine
Start set to 0.0 meaning as soon as the routine ‘trial’ is run, the text will appear and last for 1.0s duration
The text which is displayed is ‘hello world’
When we press ‘run the experiment’ button with our text - (3)
Psychopy asks us to save the experiment
You should always make a note of where it is
At this time, we will name it as ‘hello.psyexp’ (leave extensionas it is)
When running the text the output will be - (2)
It will first pop up with box saying to input participant and session to help you keep track of participant and session, gives you a randomised participant ID - assumes running psychopy as part of experiment
Press ‘okay, the screen will go blank and some text will appear for 1.0s
You can also add a a shape (polygon) in your experiment (alongside the ‘hello world’ text) in your trial routine by selecting ‘polygon’ from components:
When double clicking on polygon, these properities show up - (basic) - (3)
It is asking you to name the polygon stimuli in trial routine so far it is set as ‘polygon’
Asking for start and end time
The shape of the polygon is going to a triangle but can alter it at the drop down menu
You can also choose a hexagon in polygon properities box by
from shape menu select regular polygon with 6 sides - making it a hexagon
Now we have in our ‘trial routine’ - (2)
a ‘text’ saying hello world for 1.0
a hexagon for 1.0s
“In PsychoPy, if both text and polygon stimuli have the same duration, what determines their order of appearance?” - (2)
“The order in which stimuli are stacked in the routine determines their presentation order.
Since both text and polygon stimuli have the same duration value (1.0), the polygon stimulus will appear ontop of the text stimulus
If we run the experiment with this
polygon appears over the text and since polygon is so big, we can’t see the text
What does the appearence section of polygon properities box show you? - (5)
Fill colour: The colour with which the polygon is filled.
Border colour: The colour of the polygon’s border.
Colour space format for the specified colours.
Opacity of the stimulus: A value of 1 indicates it’s opaque, 0 means fully transparent, and 0.5 represents translucency.
Contrast of the stimulus and width of the shape.”
How to change colour of the polygon? - (3)
Next to ‘Fill Colour,’ select ‘Specify Colour.’
Choose one of Python’s colors listed on the right or changing values of RGB channels
For example, to change the polygon’s fill color to ‘salmon,’ select it from the list.”
How to change the opacity of our ‘salmon’ colour polygon to be semi-transparent? - (2)
Navigate to the ‘Appearance’ section in the Polygon Properties box.
Adjust the opacity value to be 0.5 (or 50%) to make it semi-transparent.”
When we run our code with text stimuli (still unchanged- saying ‘hello world’) and semi-transparent salmon polygon with both duration values of 1.0s, we get when we run the experiment:
Semi-transparent salmon polygon appearing over the text saying ‘hello world,’ - both polygon and text lasting for 1.0s
You can also alter more properities of the text such as in formatting section of text properities, you can - (3)
change the font of text by changing font name (e.g., Open Sans to Comic Sans)
Specificy size of text via letter height
Specificy if the text is read from left to right (set to LTR at the moment) or RTL (right to left)
You can also alter more properities of the text such as in formatting section of text properities, you can in appearence again as well as layout tab
change the colour of the text, opacity, contrast etc.. as well as its layout like position of stimulus (set at centre at position 0,0), orientation of stimulus in degrees
The order in which stimuli are stacked in the routine determines their presentation order - (3)
“How do you change the presentation order of stimuli in PsychoPy so that text appears on top of other stimuli?” (both have a duration of 1.0s)
“To change the presentation order so that text appears on top of other stimuli:
Right-click on the polygon stimulus.
Select ‘Move to Top.’”
How to remove stimuli from Psychopy for instance? - (2)
Right click on stimuli in the routine (in this example, in trial routine right click polygon)
Should be option saying remove (polygon)
How to make the polygon semi-transparent?
Fiddle with RGB colours so polygon is black ([-1,-1,-1])
Can fiddle with RGB colours to make it white by
setting to [1,1,1]
How to make polygon appear later than the text? - (2)
Set its start value to 2.0s and duration for 1.0s for example
changes the timeline of the stimuli
How can we make the polygon move? - (4)
“To make a polygon move continuously across the screen:
Go to the polygon properties and click on its ‘layout’ tab.
Find the ‘position’ entry.
Replace the X-coordinate with ‘t,’ representing the current time since the start of the experiment in seconds.
Set the menu setting to the right of the Position to ‘set every frame’ to ensure the position updates on every frame.”
What is variable ‘t’ in Psychopy? - (2)
the current time since the start of the experiment in seconds.
That number is always getting bigger.
What happens when we replce the ‘x’ coordinate of position with t? - (3)
tell Psychopy to update the position of polygon on every frame then the polygon will move across the screen:
causes the stimulus to move horizontally across the screen over time.
This is because ‘t’ represents the current time since the start of the experiment in seconds, and it continuously increases as the experiment progresses.”
How to make the polygon move to the left? - (3)
Replace the ‘x’ coordinate of the position with ‘-t’. This means that the position along the x-axis will decrease as time increases, effectively moving the polygon to the left.
Psychopy needs to update the position on every frame to ensure continuous movement.
To do this, change the menu setting next to the position property to ‘set every frame’.”
What does replacing ‘t’ in x coordinate of position with sin(t * 2 * pi
As ‘t’ time gets bigger, polygon goes up and down (along horizontal axis) like a sine wave
How to make the polygon rotate in the centre? - (5)
Make the orientation of the object dependent on time (instead of position)
Set position to 0,0 (x,y) at the centre
The orientation is set by degrees
So after 1s , move 1 degree and after 2s move 2 degrees which is not fast if orientation set to just ‘t’
So to make polygon spin faster, we can set ‘t’ to 20 so after 1 s, orientation of polygion moves by 20 degrees, after 2s orientation of polygon moves by 40 degrees etc..
Ensure orientation is updated on every frame by setting the menu setting next to the orientation parameter to ‘set every frame’.”
To produce We are now going to produce an experiment in which polygon randomly moves left or right and have to make a keypress (left or right arrow key) - (2)
we have to randomise which direction polygon goes and to do that we need to produce a little bit of code using code from component box
(we also removed the text)
You can add code to Builder projects. In fact this is the recommended way of making experiments a bit more complex in
Psychopy
When you double click the code element, it brings up:
To make polygon randomly move to the left, right or stay in the middle - (8)
Start by adding a code component to your PsychoPy Builder project. This component allows you to insert custom Python code into your experiment.
Inside the code component, navigate to the ‘Begin Routine’ tab. Here, you’ll write code that will be executed at the beginning of each trial or routine.
In the ‘Begin Routine’ tab, insert the following Python code:
direc = randint(-1, 2)
This code utilizes NumPy’s randint function to generate a random integer between -1 (inclusive) and 2 (upper range exclusive) so:
The possible values for direc will be -1, 0, or 1.
This variable direc will determine the direction of movement for the polygon.
Set the polygon’s position parameter to (t * direc, 0), where t represents time. Multiplying t by direc allows for random movement along the x-axis. If direc is -1, the polygon will move left; if direc is 0, it will stay in the middle; if direc is 1, it will move right.
To ensure that the position is updated on every frame, set the menu setting next to the position parameter to ‘set every frame.’ This ensures that the polygon’s position is continuously updated throughout the duration of the routine, allowing for dynamic movement.”
“If we set the position of the polygon to (t * direc, 0) but set it to constant - (5)
The polygon’s position is determined by ‘t’ multiplied by ‘direc’ (-1, 0, or 1).
At the beginning of the routine. so it will pick a value for ‘direc’ and then it will stay the same for the whole duration of the routine
However, the polygon remains fixed at this initial position for the entire duration of the routine.
Since the position is constant, the polygon does not update its position over time.
Therefore, it stays stuck at its initial position regardless of any changes in time (‘t’) during the routine.”
After adding code making polygon move randomly left, right or stay in middle
in image below
going to add this code in ‘end routine’
What does this piece of code do? - (6)
When builder creates a psychopy experiment it makes a data structure called ‘thisExp’
‘thisExp’ is a data structure created by PsychoPy that contains all the information about the experiment.
It keeps updating this as it goes along - and at the end it saves it to disk.
If you add in your own variables to the same structure it will save them out as well
We added the variable ‘direc’
It says save my new variable called ‘direc’ (in black) into data structure called ‘thisExp’ and call it ‘direc’ (in grey) - adds it into a column in CSV file
How to run a trial many times?
Press ‘insert Loop’ around a block of routine(s)
We can place two dots on either side of trial and ‘press’ okay - (2)
And we have produced a loop and asked psychopy to runt his trial many times
In dialog box, we can change the number of trials to ‘15’
Since we saved thisExp.addData(‘direc’, direc) we can see in CSV - (2)
date = date and moment experiment started down to millisecond, tells participant session and trials - has all info for data analysis
measures framerate - how fast screen is running (e.g., 60 Hz - updates 60 times per second) like depends temp of room, graphics card - records before each trial begins
The polygon does move randomly from trial to trial (15 trials) but its haerd to know when some trials start and end - (2)
so we add ‘pre’ and ‘post’ trial within the loop by adding a routine
In each new routine add some text: ‘get ready’ and ‘respond’ for 1.0s
You will now see when we run the experiment, there will be gap
before and after. If the ‘stimulus’ is the polygon then the gaps represent the ‘inter-stimulus interval’ or ‘ISI’.
Close to have our experiment complete but one thing is to make measurements in this case for our polygon moving left, right or centre randomly we - (6)
collecting a keypress .
We are going to ask subjects to use the arrow keys to tell us whether the stimulus moved left, right or stayed still. ‘Left arrow’, ‘Right arrow’ and ‘up arrow’.
To do this, add another component to the ‘post’ period: a keypress after trial in post routine- allowing participants to respond after stimuli stopped moving for 1.0s
Having allowined keys in this component to be ‘left’ arrow for when polygon moving left, ‘right’ so ‘right arrow’ for when polygoin moving right and ‘up’ arrow for polygon in centre
‘Force end of routine’ ensures that the routine ends immediately after the participant responds to the post-trial instruction -
then trial loop will proceed to the next iteration until the specified number of iterations (e.g., 15) is reached. After completing the specified number of iterations, the trial loop will end,
Now Pschopsy produced a directory calld ‘data’ which contains - (3)
‘log’ - produces more detailed view of experiment of 1 millisecond by 1 millisecond of whats happening on screen
‘psychdat’ produces binary representation of stimulus
‘csv’ contains all the important information of the experiment
What does the CSV contain? - (4)
1: frameRate (right at the end): Although the monitor claims to run at 60Hz, the Psychopy has double checked this and found that it is actually running at 59.78Hz.
- 2: key (in the middle): This is the response that you made. Useful!
- key_resp_rt : How long it took you to respond (after the ‘Respond!’ text came up). This is pretty precise! Perhaps it is also accurate.
- direc : This is the variable that we created! It has been logged here because we added that magical bit of code above:
thisExp.addData(‘direc’, direc)
What does this show for grating stimuli properities?
Texture within the grating stimulus has a 1-D sine wave
What does the foreground colour show for grating stimulus?
Grating stimuli will be black and white
How to produce a circular grating stimulus?
Going to change the mask to circle
How to produce 4 stripes in circular grating stimulus? - (4)
setting the Spatial Frequency parameter to 4 will indeed produce 4 stripes (4 vertical black lines and 4 white lines) within the grating stimulus.
In PsychoPy, the Spatial Frequency parameter determines the density of the stripes within the grating.
When you set it to 4, it means that within each unit of space (e.g., pixels), there will be 4 stripes.
This results in a higher frequency of stripes, making the grating appear more densely striped or patterned.
(circle now)
Placing our keyboard response withtin the trial to measure RT - (3)
using only space keypress response
Also selected ‘force end of routine’
Enabling ‘force end of routine’ means that the routine will end immediately after the participant presses the ‘space’ key - only one trial is done
Placing more one than one trial for 15 trials
“What happens in a trial routine that includes a ‘Get Ready’ text component displayed for 1.0 second, followed by a grating stimulus shown for 1.0 second, and a keyboard response allowing only ‘space’ with ‘force end of routine’ selected - , all looped for 15 trials? - (4)
“In each trial:
The ‘Get Ready’ text component appears for 1.0 second to prepare the participant.
Next, the grating stimulus is presented for 1.0 second, engaging visual perception.
Participants must respond using the ‘space’ key, triggering the ‘force end of routine’ to end the trial immediately upon response.
This entire trial sequence is repeated for a total of 15 trials, ensuring consistency and allowing for data collection across multiple iterations.” - The loop is important because we have to get lots of measurements to make a statistically-reliable estimate of the average RT.
Whats wrong with the experiment setup to measure reaction time and solution?: - (2)
“What happens in a trial routine that includes a ‘Get Ready’ text component displayed for 1.0 second, followed by a grating stimulus shown for 1.0 second, and a keyboard response allowing only ‘space’ with ‘force end of routine’ selected - , all looped for 15 trials? - (3)
In CSV under key_respon, RT is normal towards start around 200 ms but towards the end, the RT get very short around only 9 ms
What happens is individual is learning to antipcate the stimulus (anticipatory effect), the get ready! cue is taking exavtly 1 s and knowing its coming and get their finger ready and cheat on RT task
Thus, we need to get this get ready period randomised
The random wait is critical, otherwise you will just get into the habit of anticipating the stimulus and your RT will not really be a ‘reaction’.
How to randomise the ‘get ready’ cue? - (7)
Adding code element in getReady routine in which:
Set duration of ‘Get Ready’ to randTime variable
randTime equals random()*21 which generates a random floating-point number between 0 (inclusive) and 21 (exclusive) and assigns it to the variable “randTime”.
random()” generates a random floating-point number between 0 (inclusive) and 1 (exclusive).
Multiplying by 21 scales this random number to the range between 0 and 21.
. By generating a random value for “randTime”, you’re introducing variability in the timing between the grating stimulus and the onset of the “Get Ready” text.
This variability creates a variable Inter-Stimulus Interval (ISI) between the two stimuli, which can help prevent participants from predicting the onset of the “Get Ready” text
RT responses can be produced from
“What does the code ‘thisExp.addData(‘randTime’, randTime)’ mean?” - (3)
“The code ‘thisExp.addData(‘randTime’, randTime)’ adds a data entry to the experiment’s data log.
It stores the value of ‘randTime’, which represents the randomly generated Inter-Stimulus Interval (ISI) between the grating stimulus and the onset of the ‘Get Ready’ text.
It also saves a column named “randTime” to the output data file of the experiment and each row in column contain ISI value generated for each trial to CSV
Q1
In psychopy I have created a polygon. I would like its vertical position to change with time. I have set these parameters:
What’swrong?
The position is not being updated every frame. Change the dropdown on the right to ‘Every frame’
Q2
I would like to loop my trial 5 times. What is wrong with the experiment shown here?
The loop is only placed around the ‘Get ready’ routine, not the trial itself. The dots should be on the left of ‘Get ready’ and the right of ‘Trial’
Q3
This code defines a delay before the stimulus is shown. What are the minimum and maximum possible delay values?
The rand() function returns a number between 0 and 1. So the min=2, and max=6.
Q4
I need to save the value of randTime because I suspect that the delay might influence the reaction time. When I come to look at my experiment data file it has not been saved. Here is the code. What is wrong?
There needs to be something to add randTime to the experimental data structure on each trial. Possibly this statement is in the code that is run at the end of the experiment (you can’t see that but the asterix shows that something is there), but that is too late!
Q5
I have added a new variable to the code: targetSide. What values can it take? - (4)
A5:
This question was (accidentally) a ‘trick’ one. Randint is present both in random (the standard python module) and np.random (part of numpy).
Psychopy will use the numpy version: np.random.randint(x) returns a number between 0 and x-1. So the Values this variable can take are -1 and 1.
But random.randint requires two numbers random.randint(a,b) and returns a value between a and b inclusive!
We will make sure this ambiguity is not in the actual exam.
