1 Executive Summary

The aim of this report is to examine:
a) the relationship between age and reaction time
b) age and short-term memory, and
c) any correlation between reaction time and short-term memory,
in order to understand the potential impact an aging population will have for Australia.

Hypothesis:
With increased age, reaction time and short-term memory will both deteriorate.

Image from: https://knowyourmeme.com/photos/1192098-lego

Image from: https://knowyourmeme.com/photos/1192098-lego


2 Initial Data Analysis (IDA)

2.1 Context

Australia is made up of an increasingly aging population (ABS, 2016). As healthcare quality increases so too does life expectancy. Adding to this, the average number of children per household is decreasing, creating a shift towards a generally older community. To clarify statistically, “in 2017, approximately 3.8 million people (15% of Australia’s total population) were aged 65 and over” (AIHW, 2017).

Fig. 1: Proportions of population aged 0–14 years and 65 years and over, 1901–2017

Fig. 1: Proportions of population aged 0–14 years and 65 years and over, 1901–2017

There appears to be a push for areas of research to address the impact this will have, now and in the future (AGPC, 2013). We wanted to take part in this relevant area of research, as university can sometimes create a bubble of predominantly young people. We believe it’s important to have an awareness of the wider present and readiness for the future.

2.2 Definitions

Memory and reaction time are considered to be effective markers for cognitive function (Frerichs & Tuokko, 2012).

Jain AK (2012) defines reaction time as “the interval of time between presentation of stimulus and appearance of appropriate voluntary response in a subject”.

Tulving (2000; from Brady et.al. 2011) broadly defines memory as the “neurocognitive capacity to encode, store, and retrieve information”. More specifically, Atkinson and Shiffrin (1968) differentiate short-term memory - the kind of memory processing we tested - from long-term memory as information held for a finite time in a way that is easy to retrieve.

2.3 Applications of data?

There are wide applications for researching potential impacts on reaction time and memory for an aging population, including (but not limited to):
- road safety
- learning capabilities across ages
- general survival reflexes
- overall job performance
- ability to respond to emergency situations

2.4 Source

Data was sourced from…
- A Google Form https://forms.gle/GQktNwmrH1Ki5Bpg9 designed by students
- A website capable of measuring reaction time and 3 different types of short-term memory processing https://www.humanbenchmark.com/
- Researchers reached out to family members, family friends, other students, and work colleagues to act as participants.
- Data was automatically linked to and transferred from Google Form to Google Sheet (Excel Spreadsheet) https://docs.google.com/spreadsheets/d/1zh-Ft0KFWgnWBj8Y-5uRpuH_m9F1XMYqWTBzF29KCiU/edit#gid=813870

2.5 Validity and Possible Issues

Area Controlled Issues
Confounding Factors Time of day, age, gender, occupation, health/wellbeing rating, mental health condition + daily impact, hours of sleep Any substances under the influence of during testing (e.g. alcohol, caffeine, medication), wakefulness factor, average hours slept/night, any previous interaction/practice with memory and reaction time games
Anonymity? Names weren’t asked for/incldued Obvious selection bias - all participants were personally connected to both data collectors
Ages There is some spread of different age groups BUT there isn’t enough representation for: Late 20s, 30s, 40s, 60-80. You can see clustering of more data around ages of data collectors and ages of their parents
Mental Health The 2 questions pertaining to mental health were both non-compulsory/optional to answer. Asked about mental health - this is a private issue, potentially sensitive topic.
Generalisability All participants are Australian. Data was only collected for those for those living in Sydney city and surrounding suburbs. There could be a difference for those living in more rural communities, or different parts of Australia.
The website used for testing Straight forward instructions given. Doesn’t take into account those who don’t have access to computers, or those who have difficulty using technology (usually older generations). Could confound the data for those aged 50+
Anyone can take part if they have access to a computer. The website alerts its users that the reaction time test is “affected by the latency of your computer and moitor. Usign a fast computer and low latency/high frame rate monitor will improve your score.” (Human Benchmark)

Each row represents an individual participant.

  • Total number of participants = 27
  • Made up of (most importantly) differing age groups

Each column represents the results created by individual participants:

  • Time of day
  • Age
  • Gender
  • Occupation
  • Health/wellbeing rating
  • Mental health condition + daily impact
  • Hours of sleep
  • Reaction time (mms)
  • Visual Memory (points)
  • Number Memory (digits)
  • Verbal Memory (points)

3 R’s Summary of the Data

library(readxl)
project2 = read_xlsx("Data for Data1001 Project 2.xlsx")      
head(project2)
## # A tibble: 6 x 12
##   Timestamp             Age Gender Occupation `Overall Health…
##   <dttm>              <dbl> <chr>  <chr>                 <dbl>
## 1 2019-04-02 23:10:20    18 Female Universit…                5
## 2 2019-04-03 08:53:43    17 Male   Universit…                6
## 3 2019-04-03 09:35:15    26 Female Universit…                7
## 4 2019-04-03 20:14:46    49 Female home duti…                8
## 5 2019-04-04 14:32:19    18 Female Universit…                9
## 6 2019-04-04 14:53:45    58 Female Part time…                9
## # … with 7 more variables: `Mental Health Condition?` <dbl>, `Impact of
## #   Condition` <dbl>, `Hours of sleep` <chr>, `Reaction time (ms)` <dbl>,
## #   `Visual Memory (points)` <dbl>, `Number memory (digits)` <dbl>,
## #   `Verbal Memory (points)` <dbl>
dim(project2)
## [1] 27 12
class(project2)
## [1] "tbl_df"     "tbl"        "data.frame"
str(project2)
## Classes 'tbl_df', 'tbl' and 'data.frame':    27 obs. of  12 variables:
##  $ Timestamp               : POSIXct, format: "2019-04-02 23:10:20" "2019-04-03 08:53:43" ...
##  $ Age                     : num  18 17 26 49 18 58 17 15 17 57 ...
##  $ Gender                  : chr  "Female" "Male" "Female" "Female" ...
##  $ Occupation              : chr  "University student" "University Student" "University Student & Primary Teacher" "home duties" ...
##  $ Overall Health/Wellbeing: num  5 6 7 8 9 9 7 9 7 6 ...
##  $ Mental Health Condition?: num  0 0 0 0 0 0 0.5 0 0 0 ...
##  $ Impact of Condition     : num  NA NA NA NA NA NA 7 NA 4 NA ...
##  $ Hours of sleep          : chr  "2-3 hours" "7-8 hours" "7-8 hours" "5-6 hours" ...
##  $ Reaction time (ms)      : num  408 209 351 550 367 679 380 260 418 377 ...
##  $ Visual Memory (points)  : num  9 9 9 7 9 8 13 11 11 8 ...
##  $ Number memory (digits)  : num  9 8 10 10 7 9 9 8 10 6 ...
##  $ Verbal Memory (points)  : num  46 46 44 35 39 29 90 38 47 16 ...

4 Research Question 1

4.1 Is there a correlation between age and reaction time?

Overall, a correlational value of ~0.47 was found between age and reaction time. This can be described as a slight relationship.

cor(project2$`Reaction time (ms)`,project2$Age)
## [1] 0.4717033

However, the sample size was too small to see any real trends or outliers. For participants aged 15 - 24, there was a reasonably strong correlation between age and reaction, such that those around the same age had similar reaction times. However, beyond this age range participants’ results started to ‘fan out’ and become more inconsistent, with reaction times having more variation away from the regression line.

library(readxl)
project2 = read_xlsx("Data for Data1001 Project 2.xlsx")
plot(project2$`Reaction time (ms)`~project2$Age, xlab="Age (years)", ylab="Reaction Time (ms)", main="Age versus Reaction Time", pch = 19, col="darkslateblue")
abline(lm(project2$`Reaction time (ms)`~project2$Age), lty=5, col="red")

Possible reasons for some participants having faster/slower reaction times than others within the same age group may be due to each participants varying occupation/s, physical rigour and/or the time of the day that they completed the task. Results for older age groups may not have shown a great a relationship as the participants over 50 years only make up 33.3% of all subjects, with some age ranges not even being represented at all due to a lack of participants.

Age Range Number in Experiment Percent in the Experiment
<18 5 18.5%
18 - 24 6 22.2%
25 - 29 2 7.4%
30 - 34 1 3.7%
35 - 39 0 0%
40 - 44 1 3.7%
45 - 49 3 11.1%
50 - 54 1 3.7%
55 - 59 4 14.8%
60 - 64 3 11.1%
65 - 69 0 0%
70 - 74 0 0%
75 - 79 0 0%
80 - 84 1 3.7%

5 Research Question 2

5.1 Is there a correlation between age and visual memory?

When tested on visual memory, younger participants performed significantly better than older participants. R reveals a strong negative correlation between the two: (approx. -0.78)

cor(project2$Age, project2$`Visual Memory (points)`)
## [1] -0.7787682

There are clear blocks of greater visual memory in younger age groups, and lowered visual memory in older age groups. For anyone under 30, the mean number of points gained when tested is 10.17, while those 40 and over have a mean of 7.54. The highest performance is by a 17 year old, and the lowest performance is shared amongst the ages of 57, 58, 62, 63 and 81.

library(readxl)
project2 = read_xlsx("Data for Data1001 Project 2.xlsx")

plot(project2$Age,project2$`Visual Memory (points)`, xlab = "Age (years)", ylab = "Visual Memory (points)", main = "Impact of Age on Visual Memory", pch = 19, col = "purple")
abline(lm(project2$`Visual Memory (points)`~ project2$Age))

A positive side-effect of knowing the participants meant that feedback could be given directly to the data collectors, both during and immediately after testing. One participant (aged 63) commented on the speed at which original information was shown, before being asked to recall it - “Oh! It’s too fast!” So, perhaps one shortcoming of the visual memory test was that it conflated a high speed of information absorption with the ability to recall it.

While the data suggests a clear and conclusive correlation, further research would be required to substantiate any claim for causation; that, as age increases, visual memory decreases.

5.2 Is there a correlation between age & verbal memory, or age & number memory?

Both verbal memory and number memory have a weak correlation with age, as seen below with a negative correlation of approximately -0.15 for verbal and -0.29 for number.

cor(project2$Age, project2$`Verbal Memory (points)`)
## [1] -0.1497956
cor(project2$Age, project2$`Number memory (digits)`)
## [1] -0.2939556

It is important to note that while we have adopted the term ‘verbal memory’ directly from the website used for our testing for the sake of clarity, a more appropriate description for it might be categorical discrimination, since participants were shown words and asked if they were “new” or “old”.

plot(project2$Age, project2$`Verbal Memory (points)`, xlab = "Age (years)", ylab = "Verbal Memory (points)", main = "Impact of Age on Verbal Memory", pch = 19, col = "dark green")
abline(lm(project2$`Verbal Memory (points)`~ project2$Age))

plot(project2$Age, project2$`Number memory (digits)`, xlab = "Age (years)", ylab = "Number Memory (digits)", main = "Impact of Age on Number Memory", pch = 19, col = "dark red")
abline(lm(project2$`Number memory (digits)`~ project2$Age))

Summary:
There is a general downward trend overall between age and memory, but correlation isn’t strong enough to imply a direct, causal relationship.

6 Research Question 3

6.1 Is there any causation between reaction time and memory?

When testing the strength of correlation between reaction time and each different kind of memory, all have relatively weak relationships. There appears to be no correlation between the two, let alone any link to causation.
Testing correlation in R against reaction time gives a mix of results. Visual memory has a slight negative relationship (-0.35), and number memory has an even smaller negative relationship (-0.07). Verbal memory, contrastingly, has a small positive relationship (0.14).

cor(project2$`Reaction time (ms)`, project2$`Visual Memory (points)`)
## [1] -0.3451632
cor(project2$`Reaction time (ms)`, project2$`Number memory (digits)`)
## [1] -0.07342986
cor(project2$`Reaction time (ms)`, project2$`Verbal Memory (points)`)
## [1] 0.141818

Keep in mind, the lower the Reaction time (ms), the presumed better cognition, while the higher the Memory Performance, the better cognition. This doesn’t show any overall trend in any particular direction, as can be seen in the following graph. Each different kind of memory test versus reaction time has been overlayed.

par(yaxt = "n")
plot(project2$`Reaction time (ms)`,project2$`Visual Memory (points)`, xlab = "Reaction time (ms)", ylab = "Memory Performance", ylim = c(6, 13), main = "Reaction Time VS Types of Memory", pch = 21, col = "purple")
abline(lm(project2$`Visual Memory (points)`~ project2$`Reaction time (ms)`), lwd = 2, col = "purple")
par(new=T)
par(yaxt = "n")
plot(project2$`Reaction time (ms)`, project2$`Verbal Memory (points)`, xlab = " ", ylab = " ", main = " ", ylim = c(0, 90), pch = 21, col = "dark green")
abline(lm(project2$`Verbal Memory (points)`~ project2$`Reaction time (ms)`), lwd = 2, col = "dark green")
par(yaxt = "n")
par(new=T)
plot(project2$`Reaction time (ms)`, project2$`Number memory (digits)`, xlab = " ", ylab = " ", main = " ", ylim = c(6, 13), pch = 20, col = "dark red")
abline(lm(project2$`Number memory (digits)`~ project2$`Reaction time (ms)`), lwd = 2, col = "dark red")
legend(x = 700, y=13, title="Memory Types", legend=c("visual", "verbal", "number"), col=c("purple", "darkgreen", "darkred"), lty = 1, cex = 0.8)

6.2 Why isn’t there a correlation?

While both reaction time and memory do involve the occipital lobe for initial taking in of visual stimuli (UQ, 2018), the response is prompted from entirely separate areas of the brain. For reaction time, the processes required to prompt a physical response (in this case clicking), happen in the cerebellum, the back of the brain. Short-term memory, on the other hand, mostly uses the prefontal cortex (UQ, 2018).

In the context of an aging population this lends us some hope; while the speed of our body’s reactivity may worsen as we grow older, the mental capacity for memory doesn’t necessarily follow the same downward trend.


7 Conclusions

Who are some relevant potential stakeholders for this report?

  • Younger Australians - We will all grow older. It is good to be aware of what is potentially in store for us, and prepare or work towards preventative measures e.g. life-long brain-training.
  • Older Australians - Understand how your brain changes over time and be prepared for different limitations and strengths.
  • Policy-makers - Know how to appropriately tailor laws and policies surrounding aging populations for increased safety and well-being e.g. what is an appropriate retirement age? Is there an age where it becomes entirely unsafe to drive?
  • Family members - Adjust expectations of loved ones cognitive abilities at different life stages. Foster a sense of patience and understanding.

Final Conclusion:
While reaction time appears to deteriorate as age increases, short-term memory appears relatively independent of age. If an individual’s reaction time is slower, this doesn’t imply there will be issues with memory. In layman’s terms, we needn’t be too worried about the aging population of Australia and their ability to function on an everyday level in regards to their overall cognitive processing abilities.

8 References

Atkinson RC, Shiffrin RM. Human memory: a proposed system and its control processes. The Psychology of Learning and Motivation: Advances in Research and Theory. 1968. (2). pp. 89–195.

Australian Bureau of Statistics, 3101.0 - Australian Demographic Statistics, “Population By Age And Sex, Australia, States And Territories” 15/12/2016 http://www.abs.gov.au/AUSSTATS/abs@.nsf/Previousproducts/3101.0Feature%20Article1Jun%202016

Australian Bureau of Statistics. Census of Population and Housing: Reflecting Australia - Stories from the Census, 2016, 23/10/2017 https://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject/2071.0~2016~Main%20Features~Ageing%20Population~14

Australian Government, (AIHS) Australian Institute of Health and Welfare, “Older Australia at a Glance”, Sept. 10, 2018 https://www.aihw.gov.au/reports/older-people/older-australia-at-a-glance/contents/demographics-of-older-australians

Australian Government (AIHS) Australian Institute of Health and Welfare, “Deaths in Australia”, 2/07/2018, https://www.aihw.gov.au/reports/life-expectancy-death/deaths-in-australia/contents/life-expectancy

Australian Government Productivity Commission, An Ageing Australia: Preparing for the Future, November 2013 https://www.pc.gov.au/research/completed/ageing-australia

Australian Institute of Family Studies, “Population and Households” https://aifs.gov.au/facts-and-figures/population-and-households

Brady, T.F., Konkle, T., & Alvarez, G.A. (2011). A review of visual memory capacity: Beyond individual items and toward structured representations. Journal of Vision. 11 (5), 2011. pp. 1-34

Centre for Healthy Brain Ageing (CHeBA). Sydney Memory and Ageing Study. https://cheba.unsw.edu.au/research-projects/sydney-memory-and-ageing-study

Frerichs, R.J., & Tuokko, H.A. (2005). A Comparison of Methods for Measuring cognitive change in older adults. Archives of Clinical Neuropsychology. 20 (3), 2005, pp. 321-333. https://www.sciencedirect.com/science/article/pii/S0887617704001192

Jain AK. Reaction time (visual and auditory) and reflex time. Manual of Practical Physiology for MBBS. (4). 2012. p. 277-9.

QBA, Queensland Brain Institute, The University of Queensland, Australia, Where are Memories Stored?, 23/07/2018, https://qbi.uq.edu.au/brain-basics/memory/where-are-memories-stored


9 Survey

For participants: https://forms.gle/GQktNwmrH1Ki5Bpg9
For authors: https://docs.google.com/forms/d/1ApcbE7CTSGMA60reI5a4B-pR-shtDHfeQeodYiyXcoY/edit

Survey - Memory and Reaction Time

There are 2 sections to this anonymous survey. In total, it should take no longer than 8-10 minutes to complete. Part A involves 4 online tests for reaction time and memory (2-3 min). Part B is a short questionnaire reflecting on the results you’ve gained (5-8 min).

Part A

Before beginning, please ensure you are seated in a quiet environment where you won’t be disturbed for the duration of the testing. We will be using the website https://www.humanbenchmark.com for the following 4 exercises.

Click on to each of the tests in the following order and complete according to the instructions displayed on the website.
1. Reaction time https://www.humanbenchmark.com/tests/reactiontime
2. Visual Memory https://www.humanbenchmark.com/tests/memory
3. Number memory https://www.humanbenchmark.com/tests/number-memory
4. Verbal memory https://www.humanbenchmark.com/tests/verbal-memory

Once the tests have been completed, please keep the webpage open on “Dashboard” so you can see your results. You may now move on to Part B.

Part B You MUST keep your results from https://www.humanbenchmark.com open to complete this section.

This section will ask for information pertaining to mental health, but you are not obligated to disclose any information you are not comfortable with sharing.

  1. How old are you?

  2. What is your gender? Male/Female/Non-binary?

  3. What is your current occupation? E.g. university student, part-time casual retail, full-time CEO. If multiple occupations, please include both.

  4. How would you currently rate your overall health and wellbeing? On a scale of 1-10, 1 being extremely unwell, 10 being the best your health has ever been.

  5. Do you have any mental health condition/s? Yes/No/Prefer not to say/Other

  6. If yes, where on a scale of 1 to 10 would you rate the impact it has on your daily life, 1 being no noticeable impact at all, 10 being constantly noticeable impact? (this question is not compulsory)

  7. Approximately how many hours of sleep did you get last night? 0-1, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8+

  8. What was your reaction time (in ms)? Use the data you collected from https://www.humanbenchmark.com/tests/reactiontime. Please make sure that your answer is the exact number displayed from the webpage.

  9. How many points did you get for visual memory? Use the data you collected from https://www.humanbenchmark.com/tests/memory. Please make sure that your answer is the exact number displayed from the webpage.

  10. How many digits did you remember for number memory? Use the data you collected from https://www.humanbenchmark.com/tests/number-memory. Please make sure that your answer is the exact number displayed from the webpage.

  11. How many points did you get for verbal memory? Use the data you collected from https://www.humanbenchmark.com/tests/verbal-memory. Please make sure that your answer is the exact number displayed from the webpage.