Using State Space Grids to Analyze Dynamic Changes in Child and Parent Behavior

Ali Benda Author

08/02/2021 Added

79 Plays

Description

State Space Grid (SSG) coding is relatively new to the field of Psychology, and through the Learning and Development Research lab, I have had the opportunity to code complete videos second-by second to use SSG for my own research in the school year.

Searchable Transcript

Search:

[00:00:02.040]Hi everyone!
[00:00:02.790]My name is Ali Benda and I'm a senior pursuing a degree in psychology.
[00:00:07.290]I'll be doing my presentation on the use of state space grid coding to analyze
[00:00:11.490]dynamic changes in child and parent behavior.
[00:00:16.470]When interacting with their children,
[00:00:18.090]parent behavior is meant to act as a guide as to how children are supposed to act
[00:00:21.900]in different social situations. Therefore,
[00:00:24.900]parent behavior is very influential in a child's development of social skills.
[00:00:29.970]Parents' responses to their children influence their child's development of
[00:00:33.570]self-regulation skills. More specifically, warm positive responses
[00:00:38.400]encourage self-regulation skills,
[00:00:40.380]while unproductive responses tend to hinder these skills.
[00:00:44.580]Children who experienced these unproductive exchanges tend to have more issues
[00:00:48.600]with externalizing behaviors, relating back to the less developed self-regulation
[00:00:52.740]skills. Additionally,
[00:00:55.200]parental over engagement is also related to lower self-regulation skills in
[00:00:59.280]their children, as well as lower ratings of scaffolding and autonomous support.
[00:01:03.990]On the other hand,
[00:01:05.220]parents who productively engage and follow their child's lead in interactions
[00:01:09.540]promote greater attention span as well as increased behavioral inhibition.
[00:01:16.470]In the learning and development research lab,
[00:01:18.870]we are currently studying parental control and over-engagement among many other
[00:01:23.100]topics. Through the Self-regulation and Motivation In Learning Environments, or
[00:01:27.720]SMILE study, undergraduate research assistants,
[00:01:31.470]including myself are conducting two- hour-long in-person assessments that
[00:01:35.670]comprise of individual tasks and surveys completed by both parent and child,
[00:01:40.500]as well as parent-child interaction tasks.
[00:01:43.380]The specific task that I have been working on not only this summer,
[00:01:46.710]but during my time in the lab as a whole, is the homework help task.
[00:01:52.350]The homework help task examines
[00:01:54.030]how parents interact with their children while completing homework.
[00:01:57.570]The child has 10 minutes to complete six worksheets-two math and four reading.
[00:02:05.550]These are pictures of our current worksheets that our second-grade participants
[00:02:08.970]fill out with the help of their parents. As you can see,
[00:02:11.940]there are five math problems that include addition and subtraction, as well as a
[00:02:16.290]short story, followed by eight reading comprehension questions.
[00:02:22.860]In order to quantify the behaviors we observe in the homework help task,
[00:02:26.610]we code each participant video with second-by-second coding.
[00:02:30.150]This coding scheme allows us to take qualitative data,
[00:02:33.240]which in this case is parent and child behaviors, and view the data in a
[00:02:37.050]quantitative way. We code the parent and child behavior separately.
[00:02:41.190]We look for five different behaviors in parents and four in children.
[00:02:48.030]This is what our coding grid looks like.
[00:02:50.280]Child behaviors are found on the Y-axis while parent behaviors are found on the
[00:02:53.940]X-axis. The behaviors we look for in children are defiant,
[00:02:58.260]withdrawn, off tasks social exchange, and on task,
[00:03:02.500]which are coded 1, 2, 3,
[00:03:04.840]and 4 respectively. For parent behavior
[00:03:07.540]we look for intrusive, disengaged, passively engaged,
[00:03:11.770]following child's lead, and directing redirecting,
[00:03:14.890]which are coded 1, 2, 3, 4, and 5 respectively.
[00:03:22.470]This is an example of what a fully coded video looks like.
[00:03:25.620]The first column states
[00:03:26.610]the time of the video in which there was a change in behavior.
[00:03:30.000]The second column is parental behavior, and the third column is child behavior.
[00:03:34.800]Each video takes anywhere from one to two hours to code
[00:03:37.830]depending on the heterogeneity of the parent-child interaction.
[00:03:43.920]We use these codes to create our State Space Grids,
[00:03:46.980]which allow us to track dynamic changes in behavior
[00:03:49.740]temporally. Participants can only ever be in one state at a time,
[00:03:54.330]which means that the parent and child can only exhibit one combination of
[00:03:58.110]behaviors at a time. For example,
[00:04:00.840]a parent cannot be both passively engaged and following their child's lead while
[00:04:05.520]the child is on task.
[00:04:08.310]All of these possible states combined are referred to as the state space.
[00:04:13.020]Within the state space,
[00:04:14.340]our qualitative behavior that was quantified through behavioral coding can be
[00:04:18.270]tracked over time for patterns and associations between given states.
[00:04:25.080]The following grids are examples of what our second-by-second coding looks like
[00:04:28.800]when run through GridWare,
[00:04:29.610]creating a State Space Grid for us to analyze the parent-child
[00:04:33.450]interaction. In this first grid,
[00:04:35.910]we can see that a majority of the interaction involves the parent following the
[00:04:39.510]child's lead while the child was on task.
[00:04:44.130]In the second grid, we have lines to track the exact changes in behavior.
[00:04:48.210]In this grid
[00:04:48.960]we can see that while the parent had a lot of instances of following their
[00:04:52.230]child's lead while the child was on task,
[00:04:55.050]it seems the parent also spent a lot of time redirecting their child when they
[00:04:58.800]were not focused on the task, with some instances of the parent becoming
[00:05:02.430]disengaged or intrusive. Finally,
[00:05:06.960]we can see in this last grid that the parent frequently changed behavior
[00:05:10.380]throughout the course of the task while the child seems to be mostly on task.
[00:05:17.500]In order to be sure the coding we are completing is the most accurate,
[00:05:20.950]we are required to double code 20% of all homework help videos.
[00:05:25.300]These videos are double coded in order to ensure that each coder agrees on how
[00:05:29.050]to categorize specific, and typically more vague, behaviors.
[00:05:33.550]Once videos are double coded,
[00:05:35.200]we use a coding program called R to calculate reliability between different
[00:05:39.070]coders.
[00:05:40.240]In order to reach reliability codes must match 80% of the time or more for both
[00:05:45.160]parent and child codes.
[00:05:49.810]This picture shows what my R workspace looks like when I'm running reliability
[00:05:53.770]on double-coded videos. Not only does R allow us to run reliability,
[00:05:58.490]it also allows us to create Excel sheets
[00:06:00.770]to compare codes from each separate coder
[00:06:03.140]when we do not reach reliability in order to determine what behaviors we do not
[00:06:07.490]agree on in terms of categorization.
[00:06:12.740]Coming into this summer,
[00:06:13.700]the only experience I had in the LADR lab was homework help coding. Over the
[00:06:18.380]course of UCARE
[00:06:19.280]I was able to take on a more salient role in the lab in order to see how a
[00:06:23.180]longitudinal research study was maintained.
[00:06:26.360]I learned how to use both REDCap and Qualtrics in order to take notes and keep
[00:06:30.260]track of participant information,
[00:06:32.270]as well as keep track of our consent and assent forms.
[00:06:36.530]I also learned how to run child assessments,
[00:06:38.600]which has allowed me to gain a new perspective on our research process and the
[00:06:42.440]wide variety of tasks that we have parents and child complete. Additionally,
[00:06:47.420]I learned how to use the R coding program to run reliability,
[00:06:50.960]not only for coded homework help videos,
[00:06:53.360]but also for the numerous other tasks that have been coded based on behavior.
[00:06:57.650]Finally,
[00:06:58.880]throughout the process of the lab and learning all of these new things to help
[00:07:03.440]keep the lab running efficiently, I continued to code homework help videos.
[00:07:10.250]I plan on using all of the experiences from this summer to complete my Honors
[00:07:14.180]and Distinction Thesis throughout the school year. Under the advice of Dr.
[00:07:18.830]Jenna Finch and Dr.
[00:07:20.300]Chelsea Witt, I will be researching how parent-child interactions influence
[00:07:25.160]child behavior in the classroom. Using the homework help task
[00:07:28.940]in combination with teacher surveys,
[00:07:31.010]I will be examining the difference between parent control and over-engagement
[00:07:34.940]compared to autonomous support, as well as the influence of parent
[00:07:38.780]scaffolding.
[00:07:41.240]Thank you so much for listening!

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

Using State Space Grids to Analyze Dynamic Changes in Child and Parent Behavior

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments