The independent variable (sleep) has already been simulated for you (with 3 betweensubjects levels) but you will still need to label this column. One covariate, years of formal education, has also been simulated for you. However, you still need to simulate twodemographic variables (age and gender) and a dependent variable (working memoryscore). The simulation instructions are provided below and there is also a videowalk through
PSY2102 Assignment PSY2102 Assignment 1 Edith Cowan University PSY2102 Assignment Simulating Data in Jamovi Overview The independent variable (sleep) has already been simulated for you (with 3 between- subjects levels) but you will still need to label this column. One covariate, years of formal education, has also been simulated for you. However, you still need to simulate two demographic variables (age and gender) and a dependent variable (working memory score). The simulation instructions are provided below and there is also a video walkthrough. Variables Required • Independent variable Sleep (hours per night) across three levels. Provided by participants (quasi) or researcher manipulated (experimental). Means and SDs for each level decided by student. Dependent variable Working memory score. Units of measurement and range decided by student (based on a selected published measure). • Covariates/demographic characteristics Age Gender (female, male, not specified) Formal education (in years) Simulating Sample Characteristics (both designs) Age • Click a blank column heading and select “Setup”, then “New computed variable” • Label the variable “Age”. • We are going to create an age variable which will positively associate with formal education (since the two should be related). • In the formula field copy/paste: INT((NORM(25,4))+(Formal_Education_Years/2)) Note that you can modify the 25 in the above if you wish to have a slightly older or younger sample on average. Always scroll down to eyeball the list. If you are not happy with the values, simply adjust the 25 and marginally until you are satisfied. Gender • Click a blank column heading and select “Setup”, then “New computed variable”. • Label the variable “Gender”. • In the formula field copy/paste: PSY2102 Assignment Data Simulation 2 ABS(INT(NORM(0,1.2))) • This will create a variable with values of 0, 1 or 2. • Always scroll down to eyeball the list to make sure there are no values outside this range. Simply adjust the 0 or 1.2 slightly (.1 increments or decrements) until the data is sensible. • These should then be labelled in new variable. • Click a blank column heading and select “Setup”, then “New transformed variable”. • Label the variable “Gender Labels”. • Select Gender from the source drop down. • Select “Set Gender Labels” in the using transform drop down. Simulating the Dependent Variable (Quasi-Experimental Design) Before you start this procedure, you need to decide on sensible estimates of the means that you might expect for the three separate groups. That is, you need to choose 3 numbers (one for each group). There are two factors to consider when making this choice. Firstly, whether you expect there to be significant differences. If so, then you want the means to be reasonably different. If not, then make them close together. Note that the likelihood of significant differences can also be adjusted when setting the SD (see below). Secondly, you need to think about the range of possible scores for your working memory task. You want your means to be sensible (e.g., they are possible for that task). For instance, if your task has a maximum score of 20, then it would not make sense to have a mean or 20 or higher (as that would not be possible). • First, edit the labels for the independent variable to reflect the groups that you have chosen. • Click on the heading for the “Independent_Variable_Sleep” column. • Click “setup” (under the “data” tab). • Edit the labels appropriately (note that these are ordinal so label them accordingly) • E.g., the group you expect to perform the lowest on the working memory task should be the level 1 label and the group expected to perform the highest should be the level 3 label. • See below to see the highlighted values that you need to edit. PSY2102 Assignment Data Simulation 3 • Next, you will set the approximate means for each group. • Click a blank column heading and select “Setup”, then “New Transformed variable”. • Label the variable “Group Means”. • Select “Independent_Variable_Sleep” from the source drop down. • Select “Set Level Means” in the using transform drop down. • Click “Edit” and use the field below to adjust your means. • By default, they are set to 5, 10, & 15. • Change these three values to reflect the approximate means for your three groups. • See below to see the highlighted values that you need to edit. • Now it’s time to simulate the dependent variable. • Click a blank column heading and select “Setup”, then “New computed variable”. • Label the variable “Working Memory Score”. • Next we need to select an option below to paste into the formula field 1. For a continuous variable with decimal places then copy/paste: ABS(`Group Means`+ NORM(0,3)) 2. For a discrete variable without decimals, then copy/paste: PSY2102 Assignment Data Simulation 4 ABS(INT(`Group Means`+ NORM(0,3))) • The choice of decimals will depend on how you have chosen to measure working memory (continuous or discrete). • You can adjust the value of the SD (3) in the formula above to constrain or expand the spread of the distribution. • Be mindful that large SD values will likely reduce the possibility of finding significant differences. • The SD value might also need to be further adjusted to suit your desired score and this can be achieved by using an SD value that looks similar to those in the literature for the working memory task that you have chosen. • The final step requires you to eyeball the Working Memory Score data and make sure all the scores are possible (within the range of the measure you have chosen) • If you have scores exceeding the maximum possible score, then reduce the “2” in the last formula above to shrink the spread of the data. • The other option is to amend the group means make them lower. • Playing around with these two aspects of the data should bring it into line Now is a good time to “save as” to create a new file to work with for your assignment. Simulating the Dependent Variable (Experimental Design) Before you start this procedure, you need to decide on sensible estimates of means that you might expect for the three separate conditions (e.g., post-manipulation scores). That is, you need to choose 3 numbers (one for each condition). There are two factors to consider when making this choice. Firstly, whether you expect there to be significant differences. If so, then you want the means to be reasonably different. If not, then make them close together. Note that the likelihood of significant differences can also be adjusted when setting the SD (see below). Secondly, you need to think about the range of possible scores for your working memory task. You want your means to be sensible (e.g., they are possible for that task). For instance, if your task has a maximum score of 20, then it would not make sense to have a mean or 20 or higher (as that would not be possible). • First, edit the labels for the independent variable to reflect the conditions that you have chosen. • Click on the heading for the “Independent_Variable_Sleep” column. • Click “setup” (under the “data” tab). • Edit the labels appropriately (note that these are ordinal so label them accordingly) • E.g., the condition you expect to perform the lowest on the working memory task should be the level 1 label and the condition expected to perform the highest should be the level 3 label. • See below to see the highlighted values that you need to edit. PSY2102 Assignment Data Simulation 5 • Next, you will set the approximate means for each condition. • Click a blank column heading and select “Setup”, then “New Transformed variable”. • Label the variable “Condition Means”. • Select “Independent_Variable_Sleep” from the source drop down. • Select “Set Level Means” in the using transform drop down. • Click “Edit” and use the field below to adjust your means. • By default, they are set to 5, 10, & 15. • Change these three values to reflect the approximate means for your three conditions. • See below to see the highlighted values that you need to edit. • Now it’s time to simulate the dependent variable. • Click a blank column heading and select “Setup”, then “New computed variable”. • Label the variable “Working Memory Score”. • Next we need to select an option below to paste into the formula field PSY2102 Assignment Data Simulation 6 1. For a continuous variable with decimal places then copy/paste: ABS(`Condition Means`+ NORM(0,3)) 2. For a discrete variable without decimals, then copy/paste: ABS(INT(`Condition Means`+ NORM(0,3))) • The choice of decimals will depend on how you have chosen to measure working memory (continuous or discrete). • You can adjust the value of the SD (3) in the formula above to constrain or expand the spread of the distribution. • Be mindful that large SD values will likely reduce the possibility of finding significant differences. • The SD value might also need to be further adjusted to suit your desired score and this can be achieved by using an SD value that looks similar to those in the literature for the working memory task that you have chosen. • The final step requires you to eyeball the Working Memory Score data and make sure all the scores are possible (within the range of the measure you have chosen) • If you have scores exceeding the maximum possible score, then reduce the “3” in the last formula above to shrink the spread of the data. • The other option is to amend the condition means make them lower. • Playing around with these two aspects of the data should bring it into line. • Do not export the data to excel and modify it. Now is a good time to “save as” to create a new file to work with for