NetLogo User Community Models
(back to the NetLogo User Community Models)
If clicking does not initiate a download, try right clicking or control clicking and choosing "Save" or "Download".
Try It in NetLogo Web
## WHAT IS IT?
The overarching phenominon of interest is: Education equity in Low- and Middle- Income Countries (LMICs).
This experiment models teacher-pupil interactions in a classroom and allows the user to manipulate key inputs to help understand how equality of participation is either impeded or catalyzed by the interaction of student vulnerability and the teacher's response to students who answer questions (i.e., a positve teacher response or negative teacher response) and the influence of the school environment or the characteristics of school life as measured by students' perceptions of school climate.
The basic model attempts to explain the interaction between: (a) Grade 2 students' vulnerability (a function of reading ability and marginalization status - family wealthindex, disability, orphan status and older than for grade); (b) likelihood of a teacher calling on the student; (c) the teachers' response (positive or negative) to the student's answers; and (d)influence of the teacher's response on the child's willingness to raise their hand to answer again..... Ultimately, we are able to observe how equality in student participation is impacted by these interactions.
The extended model draws from real data from Grade 2 pupils in Uganda, and adds two new variables to the model:
(1) perceived school life: represented by an "chooser" in the interface that assigns schools into one of three categories - more punitive, average school life, and more postive;
(2) student exposure to adverse events - depending on the category of school life selected in the chooser interface, students in the model are assigned to one of two categories: high exposure to adverse events and low-to-average exposure or simply students who did not qualify for having high exposure to violence. Students with high exposure were students who reported experiencing multiple experiences of two out of three types of serious violence in the past school term, which are:
- multiple incidents of both physical and relational bullying;
- multiple incidents of physical punishment;
- multiple incidents of physical sexual violence
(3) the characteristics of school life (more punitive, average, more postive) also determine the distribution of vulnerability in the school/classroom. For example, in more punitive schools (<25th%ile on a measure of "perceived school climate") will have a higher percentage of students in the high vulnerability group with the opposite situation for schools measured to be "more positive (>75th %ile on a measure of "perceived school climate)
## HOW IT WORKS
The context or model environment is a classroom made up of students with varying degrees of vulnerability and who have had varying degrees of exposure to adversity - the distributions of which are given by the "nature of school life. Teachers call on a student to answer a question, the student answers, and the teacher responds either positively or negatively. The teacher's response influences the student's willingness to raise his/her hand to participate next time. Using the interface sliders and chooser, the modeler can experiment with:
(1) different probabilities of a teacher's response being positive;
(2) different impact on student willingness to raise hand/participate following a teachers response to them
(3) the modeler can choose from the chooser interface the nature of school life: more punitive; average school life; and more positive. The nature of school life determines:
--- the distribution of levels of vulnerability (high vulnerability, average, low vulnerability
--- the proportion of students who have had high exposure to adversity
Note: The three variables: nature of school life (on chooser), vulnerability level, and exposure to violence and the respective distribution of vulnerability and exposure to violence based on nature of school life are all based on real data from Grade 2 pupils in government schools in Uganda.
Shifts in equality of participation are represented by the distribution of student participation levels, which shift over time as a consequence of the teacher-pupil interactions.
Student participation levels are represented by the desk/patch color where the student is sitting, as follows: (1) light green (less vulnerable/highest participation); (2) dark cyan (average participation); (3) mustard (high vulnerability/low participation; (4) black - student reaches zero participation and quits participating.
If a pupil reaches zero participation AND is a student who has had high exposure to adversity - that student will drop out and the patch changes to light grey with no student in it.
Student exposure to adversity is represented as the color of student: dark red - high exposure; dark blue - low or average exposure (does not meet criteria for high exposure).
Note: Participation directly corresponds to the value of the variable "vulnerability".
-- Criteria for high vulnerability: non-reader, below 50th percentile in wealth index, and one or more of the following: older than expected age for grade, disability, or orphan.
-- Criteria for low vulnerability: oral reading fluence at > 13 wpm, above 50th percentile in wealth index, appropriate age for grade, and not an orphan and not with a disability.
-- Criteria for average is all the rest.
-- Teachers' likelihood of calling on a student is influenced by student vulnerability (more vulnerable students are less likely to raise their hand to answer).
-- Teachers' response (positive or negative) is also influenced by student vulnerability (teachers are more likely to respond negatively to more vulnerable children - for one reason, more vulnerable children are more likely to make a mistake.
-- The nature of the teachers'response (positive or negative) impacts the student who answers by either increasing the student's willingness to participate/raise hand next time or decreasing the student's willingness to participate/raise hand next time.
## HOW TO USE IT
The modeler observes the set up with 121 desks where one teacher and 120 students sit; The distribution of vulnerability levels can be seen in the code and depend on the nature of school life (punitive, average, positive). Very vulnerable is represented by a mustard colored desk/path; average, a dark cyan color patch, and low vulnerable, a light green patch. The monitors and plots show the distribution of these at setup and how the distribution changes over time, including students who quite participating or dropout.
The modeler can also observe students who drop out by observing the patches that have turned light grey and do not have a student in them. (Students drop out when they quit participating and are a student who has experienced high exposure to adverse events.
The modeler can manipulate (1) "probability of positive (teacher response) see slider; (2) "impact on participation" (conditioned by teacher response) - see slider; and for the extended model, the modeler can use the interface chooser to select "perceived school life" as "more punitive" "average school life" or "more positive" and see how the setup distribution of vulnerability and exposure to adversity is influenced by whether the school is punitive, average, or positive.
To run the model --- Go Button
To stop the model -- Go Button
Note the reporters given on the interace that shows the outputs:
--On Left, Totals Over time: Total counts of total ave, low, and high participaters and total who quit participating and total dropouts
--On Right, For only students that were originally "High Vulnerability" - we track countss over time of their changing status of ave, low, and high participaters and vulneerable who quit participating and vulnerale who drop out.
--On Right, for high participators, quit participating and dropout - we count the proportion of those who were originally high vulnerale of total in these high, quit, and dropout categories.
--Below the World or on the bottom you see the counts of students with high exposure to adversity and not-high exposure to adversity (based on real data and conditioned on nature of school life)
## THINGS TO NOTICE
- Observe: The Reporters and corresponding plots as you manipulate the interfaces: Counts of (a) High, (b) Low, (c) Average student participators, (d) students who quit participating and (e) dropouts
- Observe: Changes in equality of participation by observing how the patch color changes (designating high, average or low participator status given by the color of the patches/desks; light green = low vulnerability/high participators; green = average vulnerability/average participators; and mustard = high vulnerability and low participators) --- watch the environment in the big square --- slow down the speed by reducing the speed slider at the top of the interface screen; this will help you track the changes in vulnerability and participation conditioned on the teacher response and impact of the teacher response.
- Watch the plots and monitors. How do your manipulations of the sliders and choosers given on the left in green specificallly impact the students that were originally in the High Vulnerabiilty group?
- Notice (in the monitors below the world) that the number of students with high exposure to adversity will decrease because students who dropout are only among those with high exposure to adversity (e.g., rule for dropout is meeting the conditions of quit participating and student with high exposure to adveesity).
## THINGS TO TRY
- Manipulate the entry level "probability of a positive response". What happens when this is almost 100%?
- Manipulate the entry level "impact on participation". What happens when this is set to a big number or a low number? On the slider.
- What can you do (manipulate the sliders ) to achieve high participation of all students? That is, all students are high level participaters?
- Consider the impact of high exposure to adversity by comparing the number of drop-outs that occur when you choose the extremes of 'perceived school life' -- i.e., more punitive and less punitive.
- How do manipulations of the sliders and choosers given on the left in green (probability of teacher response being positive, impact of response on participation, and chooser status of 'life at school') specifically impact the students that were originally in the High Vulnerabiilty group?
## EXTENDING THE MODEL
This model can be extended in many ways, with illustrative ideas as follows.
-- Bring in a context variable: Conflict, Rural Agriculture, Urban Poor, etc
Change the entry distribution of vulnerability based on real data about marginalized groups and achievement from these different context.
-- Extend the type of range of teacher responses. For exammple, if the response is negative, provide a scale of '"negative" that depicts different levels of severity (e.g., scale from 1 - 4 representing beating, verbal reprimand, tell to sit-down, ignore & ask another) - Place this on a slider
- Scale the level of positivity. If the response is positive (based on the slider and probability of positive) a positivity scale from 1-4 might be, for example, a tangible reward, group clap for student or other praise, appreciation of trying, or joining two children to help each other.
-- Manipulate the environment that better reflects the local culture.
-- Add a shape such as hand raise and observe how often a student raises his hand to be called upon and which students get called upon.
-- Track a single child's changes in participation in each category - high, low, regular vulnerability -- over time.
## NETLOGO FEATURES
Able to make a classroom by changing the global view
Represent students and teachers visibly and use the patch color to represent changing levels of vulnerability/participation.
By using the interface chooser "percevied school life" the user can set up the model to represent the proportion of children who have had high exposure to adversity, conditioned on perceived school life.
By tracking an individual pupil and patch the user can see the differential impact of teacher resonses on their particiption.
The ability to set a model in its simplistic form and build from there.
## RELATED MODELS
In Open ABM:
## CREDITS AND REFERENCES
Do not have as of yet.