walk.game <-
  read_csv("~/R/Mini Project 3 Data Collection - DATA FORMATTED FOR R (1).csv")
walk.game=data.frame(walk.game)

Introduction

The research question which directed this study is: Does light exercise improve short term memory?

The population parameter is: the difference between short term memory measurements between a group which does not engage in light exercise and a group which does engage in light exercise. Short term memory is measured in time taken to complete a memory matching game, recorded in seconds. The population is also limited to ages 18-60 and to individuals from the Island city of Akkeshi.

There were three papers which were inspirational for the research question. First, “Acute exercise, memory, and neural activation in young adults,” by Alexis B. Slutsky-Ganesh, et. al., which looked at neural activation and memory as influenced by temporal placement of exercise. Second, “Exercise Improves Memory Acquisition and Retrieval in the Y-Maze Task: Relationship With Hippocampal Neurogenesis,” by Karin Van der Borght, et. al., which found rats perform a Y-Maze Task better with relation to exercise tasks. Third and last, “Efects of acute exercise on memory: Considerations of exercise intensity, post‑exercise recovery period and aerobic endurance,” by Paul D. Loprinzi, et. al., which found that people with higher aerobic stamina had better memories as well as memory improvements after exercise.

I suspected that the actual parameter value would be similar to those from the literature I referenced before conducting data collection. The parameter value would show that those in the group who engaged in light exercise had faster times in the memory game, indicating improved short term memory, which is supported by the inspirational literature.

Data Collection Methods

The observational units specifically were Islander adults 18-60 from Akkeshi who consented to be in this study.

I chose to limit my sample population source to one city of the Islands in order to simplify data collection. I randomly asked occupants between 18-60 to participate in my study until there were 60 who consented. I did not use a codified method in selecting which houses to check. I split the 60 participants into two groups, the group which will engage in light exercise, “walk,” and the group which would not, “no walk.” Participants were assigned to each group randomly.

The walk group was visited twice, the first time to assign the “Brisk Walk Outdoors 30 mins” task. This did take the Islanders 30 minutes to complete. As soon as the thirty minutes passed they were visit the second time, which was to assign the “Memory Game” task, and their time (in seconds) was recorded in the spreadsheet.

The no walk group was only visited once, to assign the “Memory Game,” time recorded same way as walk group.

Potential sampling error could come from lacking a codified method for random sampling, an unknown confounding variable could be present.

Descriptive Statistics

The explanatory variable for this study is light exercise, the response variable is short term memory. Light exercise is a binary categorical variable, with the walk group and the no walk group. The walk group was assigned the “Brisk Walk Outdoors 30 mins” task and the no walk group was not. Short term memory is measured with the “Memory Game” task, the results of which are reported in seconds, and is a quantitative variable.

bwplot(game ~ walk, horizontal = FALSE, data= walk.game)

Looking at the box plots alone, it does not appear that there is any association between the two variables, light exercise and memory as measured with the Islanders tasks.

favstats(walk.game$game ~ walk.game$walk)

The means between the walk and no walk group are also very similar, as are the medians. These also support the box plot showing there to be no association.

Analysis of Results

In carrying out a test of significance and a confidence interval about your population parameter(s), make sure you:

The population parameter is: the difference between short term memory measurements between a group which does not engage in light exercise and a group which does engage in light exercise. The two groups are the no walk and walk group, no light exercise and light exercise respectively.

The null hypothesis is that there is no association between light exercise and memory.

\[H_0:\mu_{walk}-\mu_{no-walk}=0\] The alternative hypothesis is that there is an association between light exercise and memory, where light exercise improves short term memory.

\[H_{alt}:\mu_{walk}-\mu_{no-walk}<0\]

In this context, a type I error would be rejecting the null when it is true, meaning that there is actually no association but the conclusion is that there is between light exercise and memory. A type II error would be not rejecting the null when it is actually false, and concluding there is no association between light exercise and memory.

The population of interest is Islander 18-60 year olds from Akkeshi. Since random sampling was used to collect participants within this age group from Akkeshi, it is reasonable to consider the sample population representative of this demographic.

stat(t.test(walk.game$game ~ walk.game$walk))
##         t 
## 0.2136168

The appropriate test statistic for the kind of variables in this study is the t-statistic. The t-statistic is 0.2136168. This is not a statistically significant value.

The validity conditions for this test are met, as the data is not strongly skewed and there are at lest 20 observations in each group.

two.sided.p.value<-pval(t.test(walk.game$game ~ walk.game$walk))
cat("\nthe one-sided p-value is", two.sided.p.value/2)
## 
## the one-sided p-value is 0.415817

The p-value is 0.41817, which means that is the probability of observing the measured data between the no-walk and walk group assuming the null is true. This leads us to not reject the null hypothesis.

Between the t-statistic and the p-value, the measured data does not demonstrate an association between light exercise and short term memory.

confint(t.test(walk.game$game ~ walk.game$walk))

The confidence interval for the data is (-6.034318, 7.474318). This means that we are 95% confident that the difference between the no-walk and walk groups’ time to complete the memory game is between -6.034318 and 7.474318. This is a fairly balanced interval (similar positive and negative values), and includes zero, which further supports not rejecting the null hypothesis.

Conclusion

The conclusion of this study is that there is not an association between light exercise and short term memory. The data did not behave as expected based on the inspirational literature which found light exercise to improve short term memory. The results of this can reasonably be generalized to 18-60 year olds from Akkeshi because that is the population demographic from which participants were randomly sampled. The same question could be investigated on real people, and would be very interesting to examine a group of college students during finals study season based on finals scores and if the exercised before studying to see if there is an association between memory more broadly and exercise.

Bibliography: references to literature mentioned in the introduction

Alexis B. Slutsky-Ganesh, Jennifer L. Etnier, Leffery D. Labban, Acute exercise, memory, and neural activation in young adults, International Journal of Psychophysiology, Volume 158, 2020, Pages 299-309, ISSN 0167-8760, (https://doi.org/10.1016/j.ijpsycho.2020.09.018)

Karin Van der Borght, Robbert Havekes, Thomas Bos, Bart J.L. Eggen, Eddy A. Van der Zee, Exercise Improves Memory Acquisition and Retrieval in the Y-Maze Task: Relationship With Hippocampal Neurogenesis, Behavioral Neuroscience, American Psychological Association, Volume 121 No. 2, 2007, Pages 324-334, ISSN 0735-7044, (DOI: 10.1037/0735-7044.121.2.324)

Paul D. Loprinzi, Marc Roig, Phillip D. Tomporowski, Amir-Homayoun Javadi, William L. Kelemen, Efects of acute exercise on memory: Considerations of exercise intensity, post‑exercise recovery period and aerobic endurance, Memory & Cognition, Volume 51, 2023, Pages 1011-1026, ISSN 1532-5946, (https://doi.org/10.3758/s13421-022-01373-4)

Presentation: Style, organization, layout, grammar, presentation of a written report, creativity. Make use to cite any work/studies you used to come up with your research question.

R code: Also, make sure that all relevant R code and output are in the body of your report.