1 Executive Summary

This report examines the respective factors affecting USYD students’ attention spans during online learning, unraveling factors impacting attention span most and whether mature students are easily distracted as younger.

Main findings reveal:

  • Studying attention span largely depends upon cultivated habits

  • Mature students alike younger students, procrastinate once awhile


2 Full Report

2.1 Initial Data Analysis (IDA)

Here is a short glimpse of our collected data:

survey = read.csv("~/USYD First Year/DATA1001/Survey for data1001 project 2.csv", header=FALSE, comment.char="#",skip = 1)
View(survey)
str(survey)
## 'data.frame':    35 obs. of  17 variables:
##  $ V1 : chr  "10/2/2020 20:32" "10/2/2020 20:50" "10/2/2020 21:21" "10/2/2020 22:29" ...
##  $ V2 : chr  "18-21" "18-21" "18-21" "18-21" ...
##  $ V3 : chr  "1st Year" "1st Year" "1st Year" "4th Year" ...
##  $ V4 : chr  "Female" "Female" "Female" "Male" ...
##  $ V5 : chr  "Perhaps a little/ somewhat or slightly" "Perhaps a little/ somewhat or slightly" "Perhaps a little/ somewhat or slightly" "Not really" ...
##  $ V6 : chr  "Browsing and sending memes around, Netflix, Youtube and other entertainment sites to watch movies, films and sh"| __truncated__ "Looking at viral videos (9gag, facebook, reddit, tiktok etc.), Call up friends, text friends to meet up etc." "Looking at viral videos (9gag, facebook, reddit, tiktok etc.), Call up friends, text friends to meet up etc." "Netflix, Youtube and other entertainment sites to watch movies, films and short clips" ...
##  $ V7 : num  5.5 3.5 3.5 3.5 5.5 3.5 3.5 3.5 8.5 3.5 ...
##  $ V8 : int  4 5 5 4 3 4 3 5 4 3 ...
##  $ V9 : num  2.2 4 4 0.3 2.5 1 1 1 0.1 2 ...
##  $ V10: chr  "More stressed out" "Same as face to face" "Same as face to face" "Enjoyable" ...
##  $ V11: num  2 1.5 1.5 2.5 2 2 1.5 2.5 2.5 2 ...
##  $ V12: int  4 6 6 6 8 6 6 1 4 7 ...
##  $ V13: int  7 6 6 7 5 6 4 5 3 4 ...
##  $ V14: chr  "3-4 apps" "4+ apps" "4+ apps" "0 apps" ...
##  $ V15: num  5.5 3.5 3.5 3.5 5.5 1.5 1.5 1.5 0 3.5 ...
##  $ V16: chr  "Pleasure first obviously :)" "Nah, work first leaves more fun later  ;)" "Nah, work first leaves more fun later  ;)" "Pleasure first obviously :)" ...
##  $ V17: chr  "loads, feel unmotivated - take timely breaks so mind can refresh" "Shut down electronic devices except laptop and clear off my desk" "Shut down electronic devices except laptop and clear off my desk" "A lot, Need to remove distractions. " ...
## R's classification of variables
require(dplyr)
survey <- mutate_if(survey,is.character,as.factor)

require(varhandle)
Age <- unfactor(survey$V2)
Age.range = as.numeric(as.character(Age))
Extra.studying =(survey$V15)


Length.of.lectures = survey$V11

Productivity.Apps.used = survey$V14

Total.procrastination.time = survey$V7

Gender = survey$V4
Time.zone.matters = survey$V5
Distractor.tasks = survey$V6

Summary:

  • The data is retrieved from the survey our group designed on October 2020: https://forms.gle/jRpp87JRvJasY4Wx7
  • The data is valid because results were strictly gathered from credible sources, notably from our target audience: USYD students, as the survey was posted on the Ed and Reddit Forum both exclusively used by USYD students.
  • Possible issues include variable classifications, constraints in values which caused us to make minor adjustments to correct the dataset (i.e changing 1-2 hrs to 1.5, separating columns to ease data mining and visualisation for morn, aft lecture attention spans etc.)
  • Rows represent factors affecting attention spans, columns represent fellow USYD students’ thoughts on how online learning & other factors impacted their attention spans

Limitations:

  1. Mostly based off self-report stats & sample size was too small to yield highly significant trends.

  2. Confounders like mood, motivation, habits, stress varied different times and days participants fill the survey, so it is unsuitable to assume long-term trajectory patterns are continuous due to subjectivity.

  • Future adjustments on the survey could try to address these issues and limitations by investigating the effect of those specific confounders on attention spans.


2.2 Which defining factor affects USYD students’ attention span the most since transitioning online?

require(ggplot2)
require(tidyverse)

  survey %>% 
  rename("distractors" = "V6",
         "gender" = "V4") %>% 
  mutate(distractors = str_replace(distractors, "Looking at viral videos \\(9gag, facebook, reddit, tiktok etc.\\)", "Viral videos"),
         distractors = str_replace(distractors, "Browsing on instagram - insta-feed, editing and adding pics \\(exploring filters\\)", "Instagram browsing"),
         distractors = str_replace(distractors, "Netflix, Youtube and other entertainment sites to watch movies, films and short clips", "Entertainment sites"),
         distractors = str_replace(distractors, "Call up friends, text friends to meet up etc.", "Call and text friends")) %>% 
  separate(col = distractors, sep = ",", into = paste0("choice_", 1:10), fill = "right") %>% 
  pivot_longer(cols = paste0("choice_", 1:10), names_to = "choice_number", values_to = "choice_value") %>% 
  filter(!is.na(choice_value)) %>% 
  select(-choice_number) %>% 
  mutate(choice_value = str_trim(choice_value)) %>% 
  ggplot(mapping = aes(x = choice_value, fill = gender)) +
  geom_bar(position = position_dodge2(preserve = "single")) +
  coord_flip() +  labs(title ="Gender preferences for various distractions during study", x = "Study distractions", y = "Popularity of engaging in those different distractions")

Proportionally, females have comparatively more distraction preferences so are more prone to distractions than males, who only exceeded females’ in scrolling new news updates as a distraction.

Distracted.after.procrastinating = survey$V8
p4 = ggplot(data = survey, aes (x=(Time.zone.matters), y=Distracted.after.procrastinating, color = Time.zone.matters))+ geom_boxplot()+ xlab("Did time zone matters impact study?") + ylab("Level of distraction rate after just glimpsing socials & distractors")+ theme_light() + labs(title ="Were students more distracted due to timezone differences?")
p4

Those most affected by timezone differences managed to somewhat focus more during study compared to those only impacted a little or not really impacted as seen from the slightly lower median of the last boxplot.

barplot(table(Productivity.Apps.used), col = "paleturquoise2", xlab = "Amount of productivity apps used whilst studying online", ylab = "Popularity of students using those apps", main = "How popular used are productivity apps to aid online 
study, even those with timezone problems?")

Majority of students didn’t use productivity apps to aid study, even when timezone problems impacted them heavily. Progressively less students used more productivity apps to aid study. 1/3 students using 1-2 apps diminishing to almost 2/35 using 4+ apps.

morn.attention = as.factor(survey$V12)
aft.attention = as.factor(survey$V13)
mornattspan <- table(morn.attention)
aftattspan <- table(aft.attention)
test <- rbind(mornattspan,aftattspan)
barplot(test,beside = T, col=c("lightblue","violet"),legend= c("attention span-morn lect","attention span-aft lect"), args.legend = list(x="topleft"),ylab = "collective rankings for morn,aft lecture attention spans", xlab = "Rating of morn, aft lectures attention spans with 1 = Worst, 8 = Best", main = "Attention spans in lectures during online study") 
abline(v=10.35,col = "blue")
abline (v=14.75,col = "purple")

Mean attention span for morning lectures:

mean(mornattspan) 
## [1] 4.375

Mean attention span for afternoon lectures:

mean(aftattspan)
## [1] 5.833333

Clearly, afternoon is much preferred for studying attentively as the mean afternoon attention spans are higher than morning.

p = ggplot(survey, aes(Total.procrastination.time,Length.of.lectures))+ geom_jitter(aes(fill = Extra.studying), pch = 23, cex=4, lwd = 2) + scale_fill_gradient (low = "red", high = "yellow") + labs(title = "Exploring correlation between lecture lengths vs total procrastination times", x= "Total time spent procrastinating (hrs)", y= "Lecture length (hrs)")
p

Data is widespread on the scatter plot above but more concentrated in the middle meaning slightly longer lectures extends procrastination and more extra studying later. Yet, small outliers reveal several students spending little to no time studying outside of class, so correlations are likely weak between those variables.

Apparently, lecture timings and students’ cultivated habits are the most defining factors affecting attention spans encompassing students’ sleeping schedule, study motivation and self-discipline habits.

2.3 Do mature uni students have better study habits compared to early years or are just as easily distracted during scheduled study?

require(dplyr)

Academic.year = survey$V3
Pleasurable.activities = survey$V16
Fun.or.work.first.according.to.different.year.groups = table (Pleasurable.activities, Academic.year)

mosaicplot(Fun.or.work.first.according.to.different.year.groups, main = "Pleasure or work first according to different year groups",
           sub = "Preferences", 
           xlab = "color",
           ylab = "Maturity",
           las = 1, 
           color = c("skyblue2","chocolate","red","green","purple"), 
           border = "black", off = 15
           )

1st and surprisingly 3rd years prioritize pleasure before studying, whilst second years proportionately prefer working first. Too little data was collected for 4th and above years to make definite judgements on them, but maturity still does not negate procrastination occurring.

require(plotly)
p2 = plot_ly(survey, x= ~Age, y= ~Extra.studying, type = 'box') 

p2 %>% layout(title = "Do more mature students spend more time self-studying outside class?", xaxis = list(title = "Age"), yaxis = list(title= "Time spent studying outside of class (hrs)")
              )

Age difference doesn’t really matter when choosing to study more or less outside class, personal preference matters more.

Normal.study.before.procrastination = survey$V9
plot(Normal.study.before.procrastination, Total.procrastination.time, xlab = "Time able to focus before distracted (hrs)", ylab = "Total time procrastinated (hrs)", main = "Scatter plot of total procrastinated time vs time before distracted")

#Scatter plot with green regression line 
Linear.model = lm(Total.procrastination.time ~ Normal.study.before.procrastination)
abline (Linear.model, col = "seagreen3")
points(mean(Normal.study.before.procrastination), mean(Total.procrastination.time), col = "blue", pch = "+")
legend ("topright", c("Point of averages(centre)"), col ="blue", pch ="+")

Regression line (green):

Linear.model$coeff
##                         (Intercept) Normal.study.before.procrastination 
##                           4.2877353                          -0.1498486
n = length(Normal.study.before.procrastination)

Correlation co-efficent:

cor(Normal.study.before.procrastination, Total.procrastination.time) * (n-1)/n
## [1] -0.08288179

A weak, inverse correlation is oddly indicated by the regression line and correlation coefficient on the scatter plot above.

res = Total.procrastination.time - Linear.model$fitted.values 
plot(Normal.study.before.procrastination, Linear.model$residuals, xlab ="Focus before distracted (hrs)", ylab = "Total time procrastinated (hrs)")
abline (h=0, col = "coral")

The residual plot is also unevenly spread. Thus, further predictions won’t be appropriate since the data is heteroscedastic.

Overall, mature students are no different to younger, prone to procrastination too and weirdly, low correlation between procrastination time and focus before distractions may indicate their complex relationship (non-linear) - impacted by many confounders as well as our tiny data sampling size.


3 Research & References

Phones and other electrical devices weaken our focus on work and hijack our attention spans, thus diverting our energy devoted to tasks, especially boring and complicated ones (Budd, 2017). As such, encouraging multitask habits that disrupt study and work performance (May & Elder, 2018).

Budd, K. (2017, November 27). Attention Spans, Focus Affected By Smartphone Use. Retrieved from AARP website: https://www.aarp.org/health/brain-health/info-2017/mental-focus-smartphone-use.html

May, K. E., & Elder, A. D. (2018). Efficient, helpful, or distracting? A literature review of media multitasking in relation to academic performance. International Journal of Educational Technology in Higher Education, 15(1). https://doi.org/10.1186/s41239-018-0096-z


4 Acknowledgement Section

Group members: Elaine Chen (500013544), Claudia Lam (500305173) and Alister Liang (490054448)

Group meetings:

1st & 2nd meeting -> 2nd Oct (Friday) at 3:20 - 5:50pm and 10:55 - 12:05 am {Total: 4hrs30min}

Here, Elaine and Claudia discussed and finally decided on a research topic to base our survey on, making progressive improvements to our survey questions and determining which ones were worth keeping.

3rd meeting -> 3rd Oct (Saturday) at 3:40 - 4:40pm {Total: 1hr}

Elaine and Claudia further discussed our current progress and targets to achieve according to the rubric (i.e setting deadlines after each meeting). This time, we managed to create 2 integrated questions to answer for the report. Suddenly in the middle of the meeting, Alister messaged Elaine on Wechat about wanting to join the group, so we decided to let him join although we’ve already finished most of the survey work and only had to collect data for analysis.

4th meeting -> 6th October (Tuesday) at 4:58 - 6:50pm {Total: ~2 hrs)

5th meeting -> 9th October (Friday) at 6:56 - 9:19 pm (Total : ~2:30 hrs)

6th meeting -> 17 October (Saturday) at 5:00 - 7:20 pm {Total:~2:20 hrs}

7th meeting -> 20th October (Tuesday) at 5:40 - 7:00pm {Total: ~2:20hrs}

8th meeting -> 21st October (Wednesday) at 5:00 - 8:40pm {Total: ~3:40hrs}

9th meeting -> 22nd October (Thursday) at 5:00 - 7:38pm {Total:~2:40hrs}



Contributions:

Claudia:

  • Survey work with Elaine (brainstorm ideas, wrote questions)

  • Tidyied & troubleshooted codes

  • Video editing, completed own part on video too

  • 4 solid plots (+ 1 of Ali’s graph - gender & distractor task)

  • IDA, proofread all aspects + added own flair whilst shortening words to reach word limits

  • Research work finding & analysing articles

Elaine:

  • Survey work with Claudia (refined questions, sent surveys out)

  • 3 solid graphs (with Ali’s timezone graph)

  • Video parts

  • Executive summary

  • Summarise conclusions for our research qs

Alister: (came in late - no survey work completed)

  • Video part

  • 2 graph attempts, ran into many errors and gave up, Elaine and Claudia had to help carry the load and finished them for him