91265 experiments assessment (2016) problem · pakuranga college page 4 of 12 2016, as91265 (v2)...
TRANSCRIPT
Pakuranga College Page 1 of 12 2016, AS91265 (v2) Experiments
91265 Experiments Assessment (2016)
Problem:
Question: Is there a difference in a person’s processing time of a task when a cognitive, semantic
interference is introduced?
Purpose: To investigate a person’s processing rate of a reading task when cognitive interference is
demonstrated by the use of Stroop Test.
RESEARCH:
The Stroop Test effect is a widely known test that is able to test out a person’s “reaction time of a task”
when a cognitive, semantic interference is introduced.
Based on this website: http://www.whatispsychology.biz/about-stroop-effect-definition, it’s been said that for
this Stroop Test, when people read a word, we process the “color of each word” and also “the meaning of
each word”. So if the words’ colour and meaning match, subjects are able to process and read the word
accurately and “rapidly”. But if the word’s colour and meaning don’t match, then processing and reading the
non matching coloured word accurately, will take a longer time. This is because, “experience has taught us
to attach more significance to the meaning of words rather than the colors they are written in”. So reading
the words is naturally automatic for a lot of people. Thus, when people are asked to do the opposite and
read the colour of the word instead, a cognitive, semantic “interference occurs”.
In this website:https://faculty.washington.edu/chudler/words.html, it says that “the interference between the
different information (what the words say and the color of the words) your brain receives causes a problem.” Hence, it
would take a bit more time to process and read the colour of the non matching coloured word accurately as
we’re not familiar in doing so, compared to us reading the word. Consequently, lowering our reaction times.
Due to this, I think that participants for this experiment will have a faster reaction/processing time when
reading the matching coloured words list than naming the colour from the non-matching coloured word list
as “both the stimuli (colour and meaning) in the test are compatible” and no interference is made. Thus,
subject can automatically name the correct background colour of the listed words, hence, resulting in a
higher processing time.
Plan:
I will design a plan, as that way, I will clearly know what I’m going to do in my experiment and thus minimise
the errors in my experiment process. Hence, I will be able to hopefully produce accurate and reliable results
that will link with the research I have done and thus, answer my question and successfully complete my
purpose.
This experiment is a paired experiment as I will be comparing two sets of data measurements-matching
coloured words (RED) and non matching coloured words (RED)-from the subjects’ processing time of the
reading task.
The treatment variable is the loud, accurate reading from the list of non-matching words we’ll create as
this is where the cognitive, semantic interference is introduced via the Stroop test effect. This will be done
by letting an individual in a group, read the colour of the word from a list of non-matching coloured words,
Pakuranga College Page 2 of 12 2016, AS91265 (v2) Experiments
aloud and as fast as they could while we time them with a stopwatch. As this way, we would be able to see
if cognitive, semantic interference does affect a person’s processing time of a naming task based on their
reading time difference.
The response variable is the time it takes in seconds, for the subject to finish reading correctly the colour
of the words from 2 lists of coloured words that is based on the stroop test. They will be reading the colour
of the words aloud as that way, we will be able to observe their processing rate of the task. This will be
done by timing them, via stopwatch, when they start reading aloud and stopping the time after they have
correctly read the whole list provided.
The controlled variables are
● The font size of coloured words in the list. As we will be creating the list of coloured words in an
online document (google docs), we need to make sure that all our subjects can clearly read and see
the listed words and their colour during our experiment. So in an A4 paper, we would make sure
that all words will be in the “30” sized font and printed in colour.
● We will also control the font of the words in our list as there are some font styles that make it harder
to read the words typed/ written. So to minimize this effect, we will type up the words on google
docs with “Arial” as the font style, as this font can be clearly read by everyone.
● We will control the amount of words per list, as having different amount of words would also alter the
results of this experiments as a shorter amount of words would increase the rate of the subjects’
processing rate despite the semantic, cognitive interference introduced. We will control this by only
using 18 coloured words per list. And these words would be the same except with different coloured
backgrounds.
● It’s a fact that some people have different eye sights than normal. So we will also make sure that all
our subjects are not colour blind or don’t have any problems with coloured words, because if they
do, this might cause outliers or extreme values in our experiment due to their different eye sights.
So to minimize this, before subjects start the experiment, we will ask “Are you colour blind?”, “Can
you see colour normally”? And if they answer yes to these questions, then we won’t use them as
part of the investigation. And if they have no problems, then we will continue on with the experiment.
● As subjects would be doing the experiment twice- one with the treatment introduced and one as the
placebo, by the time they do their second attempt, subject may already be familiar with some of the
words on the list, thus, increasing their processing time in their second attempt despite the cognitive
semantic interference. So to minimize this, we will consistently make all the subjects walk to a
certain distance inside the classroom (from current place to the next nearest desk), before making
them come back to us to do the second testing with the treatment introduced. That way, they
wouldn't name the words the have already memorized from the first test.
·
Equipment used:
· 3 stopwatches
· 3 Tables to record data
· 3 lists of 18 coloured words that matches with the word meaning and
background colour (placebo).
· Eg. RED
· 3 lists of 18 coloured words that don’t match with the word meaning and
background colour (treatment). Eg. RED
Pakuranga College Page 3 of 12 2016, AS91265 (v2) Experiments
(The matching and unmatching coloured list created →)
Experimental Plan:
1. Create list of coloured words based on Stroop test via google docs that way we would be able to
perfrom the experiment.
2. Print list in colour and print 6 copies of it as 3 are needed for each subject and other 3 for testers.
3. Ask subjects questions before starting experiment: “Are you colour blind?” “Can you not see colours
clearly?”-If answer is no, continue on with experiment. If answer is yes, don't do experiment and
thank them for wanting to participate. As that way, all our data gained would be reliable.
4. Individually, tell subjects what to do for experiment: “Read the colour of the words. As fast as you
can while we time you with a stopwatch”. As to further remind participants of their task as it is quite
confusing.
5. Give subject the paper containing the list of matched coloured words and make sure it faces down
until experiment is ready to start. So that they won’t start reading before the time has started.
6. Say “Ready, Set Go”, and start stopwatch after “Go”. So that reading of the words and pressing of
the “Start” button on stopwatch would be insync.
7. Stop stopwatch after subject named all colour background in the list accurately. So that
measurement would be accurate.
8. Immediately record data in a table as to not forget the time and thus lose the data gained.
9. Make subject walk to the front of the room at an even pace and back to tester again to do the
experiment with the treatment introduced.
10. Repeats steps 4-8 but instead of the matching
coloured word’s list, give them the non-matching
coloured words. So that, method of experiment would
all be the same.
Data Collection & Recording:
Once subject has finished with the reading task, I will
immediately record the time shown by the stopwatch that I
was using, to my data table shown here.➡
For me, I will use a class of 20-30 Year 10 and Year 9
students to test the experiment. They will then be already
randomly separated into groups of 2-3 students by their
teacher(Randomisation). So conducting the experiment and
recording their results into the data recording table would be
easier due to the smaller group and also, due to the
randomisation of the students in the groups, subject biases
would be minimised.
Replication:
To improve the reliability and accuracy of my results,
1. I would repeat my tests again on a different group of students.
2. I would repeat my tests again on a larger group of students.
Also as I can only do the experiment during the 1 hour period, I wouldn’t be able to replicate this Stroop
Test to my 25-30 subjects. But if I did have time, I would do at least 2 replicated tests, that way, we would
be able to get more accurate results. Although I wouldn't do more than 2 replicated tests as my subjects will
Pakuranga College Page 4 of 12 2016, AS91265 (v2) Experiments
get use to the cognitive semantic interference of the the non matching coloured words. Thus, also
ultimately hindering the reliability of my results.
We will also do a pilot run before our actual data collection to other groups in our class as that way, we will be able to
see and fix the problems of our experiment that we’ve encountered during the pilot run. Thus, limiting the errors we
will make on the actual day of collecting our data results, hence improving the reliability of results gained.
Data: I have checked through the data and have not found any doubtable results. Although there were a
few outliers or extreme values found but these had reasons behind it as mentioned in the notes we made during or data collecting period.
# of subjects
Gender
Matched Words ( secs)
Unmatched Words (secs)
Difference (secs)
1 Girl 6 17 11
2 Boy 6 21 15
3 Girl 8 13 5
4 Boy 7 13 6
5 Girl 7 16 9
6 Boy 6 18 12
7 Girl 6 15 9
8 Girl 6 34 28
9 Girl 6 16 10
10 Boy 8 16 8
11 Boy 12 37 25
12 Boy 10 31 21
13 Girl 8 15 7
14 Boy 8 17 9
15 Girl 5 17 12
16 Boy 7 20 13
17 Girl 11 18 7
18 Girl 6 18 12
19 Girl 13 18 5
20 Boy 6 13 7
21 Girl 5 14 9
22 Girl 6 18 12
23 Boy 6 14 8
24 Girl 9 13 4
25 Boy 7 16 9
26 Girl 8 14 6
Pakuranga College Page 5 of 12 2016, AS91265 (v2) Experiments
27 Boy 7 24 17
28 Boy 7 17 10
29 Girl 4 11 7
30 Boy 6 14 8
Analysis:
DOT BOX PLOT GRAPH OF DIFFERENCE:
Centre/Median:
● I notice that my mean (10.7) and median (9) are both positive as shown by the summary statistics
next to the graph. This then means that the average time of reading the unmatched words (the
treatment) have increased overall.
● I notice that there’s quite a bit of difference in the mean and median values as there’s almost a 2
second difference which is shown via summary statistics where the mean is 10.7seconds and
median is 9 seconds. This then means that the distribution of the data is almost symmetrical and
also quite reliable. But may need further investigation for certainty.
● I also notice that the mean and median values are far from the 0 value as proven in the graph where
0 seconds is not even shown in the horizontal axis. Thus, furthermore suggesting that there is
definitely an increase in time after treatment was introduced to the patient.
Spread(range, IQR,sd):
● I notice that the IQR is in between 7 seconds(lower quartile) and 12 seconds (upper quartile) as
shown in the graph. This then means that the middle 50% of the participants’ processing time
difference ranged from 7 to 12 seconds.
Pakuranga College Page 6 of 12 2016, AS91265 (v2) Experiments
● I noticed that the minimum value is 4 seconds as shown in my graph’s summary statistics. This then
further supports the idea that all of the subjects in this experiment’s reading time have all increased
after the introduced treatment (reading unmatched words).
Shape:
● I noticed that the data distribution is almost symmetrical and is normal (bell-shaped) as there's only
one peak at 9 seconds.
● But despite that, the data has a skewness to the right due to the values: 21, 25 and 28 seconds
(maximum value) which is shown by the long tail from the graph. This then means that the if it
wasn’t for the 3 extreme values of 21, 25 and 28 seconds, the graph would be resemble symmetry
in the data distribution.
● I can see that there’s a cluster around 7 seconds that's also quite based around the 9 seconds
median value. Thus, this might be suggesting that if this experiment was repeated with more than
30 participants, the median of that experiment would be around 7-9 seconds.
Unusual Values:
● I can see that there are unusual values which have a difference of 21, 25 and 28 seconds.
According to my obtained data, these values were from a Girl (whose initial reading time was 6 and
final reading time-after treatment was introduced-was 34 seconds), a Boy (initial reading time was
12 and reading time of mismatching words was 37 seconds) and another Boy (whose initial reading
time was 10 and reading time of mismatched words was 31 seconds). One of these boys (25
second difference) were still new in learning the English language. So this most probably did have
an effect with his results, both matching (12 secs) and mismatched words’ reading time(37 secs).
For the other extreme values, these could very well be outliers due to unreliable methods of timing
their reading time.
Overall-in relation to 0 difference:
● As the median difference is 9 seconds with all results above 4 seconds difference, I can clearly see
that every participant in this experiment (100%) had all increased their processing time of a reading
task when a cognitive interference (mismatched words). This is also clearly evident when the ‘0’ in
the scale, didn't even appear on the horizontal axis of my graph-Is this normal for investigations with
the Stroop Test? If not, then could this be happening due to my small number of participants?-
Either way, I can say that with this graph, the cognitive interference introduced (mismatched words)
did have a positive effect in the participant’s processing time of a reading task.
PAIRED LINK GRAPH:
Pakuranga College Page 7 of 12 2016, AS91265 (v2) Experiments
General Trend:
● I can clearly notice that there is an increase in all of the participants’ processing rate in a reading
task when the cognitive interference (unmatched words) or treatment is introduced as 100% of the
arrows in the graph have all moved to the left. Although subjects’ increased processing rate varied
as some arrows are almost linear, while other values are angled more towards the right like the
boy’s arrowed line that moved from 12 seconds to 37 seconds. With this, although the data results
varied, I believe that my introduced treatment worked in this experiment.
Number or % of results increasing vs decreasing:
● All participants in this experiment (100%) have increased after treatment was introduced, with an
average of 9 seconds difference.
Overall spread of results – before vs after:
● I notice that there are a lot of clusters when matching words were read which is shown in the
‘Matched Words (secs)” line around 6 seconds. This then means that a lot of subjects’ processing
rate in reading the matching words were around 6-7 seconds. This is further supported from the
data table where the mode of matching words was 6 seconds.
● Although there were quite a few clusters before the introduced treatment, as shown in the
‘Unmatched Words (secs.)’ line, the data distribution is more spread out since the range of the
unmatched words is 26 seconds (Max: 37-11 (Min)) while matched words’ range was 14 seconds
(Max=18, Min=4). Meaning that there's more variability in the data after introduced treatment. Thus
showing, that overall data distribution is not quite near the middle of the distribution (median).
Pakuranga College Page 8 of 12 2016, AS91265 (v2) Experiments
● The standard deviation calculation provided in the summary statistics for non matching words
(6.1191) is 4.0735 more than the standard deviation for matching words (2.0457). Thus furthermore
proving the greater spreadness and variability in the non matching words distribution than the data
distribution in the matching words.
Conclusion:
For this investigation, I asked the question: “Is there a difference in a person’s processing time of a task
when a cognitive interference is introduced?” Based on my finding and analysis for this experiments, I
would say that yes, there is a difference.
Looking back to my data collection, and analysis of my graphs and summary statistics, it is clearly apparent
for this experiment, that the subjects’ processing rate in the reading task, have all increased when they
were reading the mismatching words (the cognitive interference introduced) after reading the matching
words (used as placebo). As all the arrows have clearly moved to higher times when the treatment was
introduced and also the data had a minimum value of 4 seconds difference-minimum increased value time.
This then means that for this investigation, all my participants were affected by the cognitive semantic
interference from the nonmatching words based on the Stroop test, thus proving our treatment to be
successful as there was quite a big positive difference in their processing time when treatment was
introduced.
My results and answers for this investigation question agrees and is also supported in my research as the
webpage states that “The results of the tests showed that most persons tend to say the color of each word for
columns 1(matching coloured words) with little problem. Despite the clear instructions however, most persons
struggle to say the color of each word in column 3. This difficulty occurs due to semantic interference.”
http://www.whatispsychology.biz/about-stroop-effect-definition
From this, I believe that the Stroop test-reading of non matching words-(cognitive interference)-did increase
people’s processing time of a task.
Pakuranga College Page 9 of 12 2016, AS91265 (v2) Experiments
Although I know that this might not be the case for everyone as there are other factors that would also
contribute to the data results such as non-fluent English speakers (where their testing times have been
slower than average), and past experiences with the Stroop test. Thinking back to when we conducted our
experiment, this particular source of variation (past experience with the Stroop Test) was something we
encountered when Student14 from my data, was not only a bit colourblind (although he could do the test
just fine) but also have done the Stroop Test before. Due to his past experience, he was one of the fastest
participants who were able accurately read the non matching words list and didn't come out as an outlier in
the graph. Because of that, in the end, we decided to just use his data for this experiment too as we
needed every single participants in the class to at least get 30 data results. But if we had more time to do
the experiment, I would’ve tested out at least 5 more junior students to replace his and the non-fluent
English speakers’ data, that way our results would have been more reliable. (Controlled Variables in my
‘Plan’ were all successfully controlled and thus, have not affected the results we obtained aside from
Student 14 as no problems arose during our attempts of controlling them during the experiment).
I also hypothesized that “participants for this experiment will have a faster reaction/processing time when
reading the matching coloured words list than naming the colour from the non-matching coloured word list”.
Based on my data findings and table, I can say that I was right in my hypothesis as all of the participants for
this experiment all had faster times reading the matching coloured words’ list, compared to the non
matching coloured list. This is also supported in my average mean and median calculations from my arrow
graph where matching coloured words time’ mean (7.2333 seconds) and median (7secs) were considerably
lower than the mean (17.933secss) and median (16.5 secs) time average for non matching coloured words.
With all this, I can say that I am quite confident in my results for this investigation (my data and conclusion)
as it not only supports my hypothesis but also my research of the Stroop Test effect. But to further increase
my confidence in the reliability of my investigation, I would perform this investigation again to a larger group
of people to reduce the big variation made by the 30 individual people from my group of participants when I
introduce the treatment variable. (Variation shown in my Arrow graph and mentioned in ‘Analysis’).
Evaluation:
When I think about the method we conducted during our experiment, it wasn’t exactly reliable as we made
room for inaccuracies with our inconsistencies. This is because as we performed the experiment, my whole
group (3 people) individually tested our participants and individually recorded the time unto our own copies
of the recording table (table was the same for everyone) and we haven't discussed and clarified how we will
individually do the experiment-from what we say to the participants all the way to our recordation of the time
results. Because of this, our method of collecting the data weren’t all consistent. For example, both me and
another member included the milliseconds when recording the time in our printed out table while the other
member did not. And also what we said to remind our subjects about their task were also all different. This
would have negatively impacted the timing results we obtained as we may have unknowingly, not been
clear to other participants while we were crystal clear in our instructions and reminders to another
participant. Thus, our results may not be truly accurate as some may not have understood at the start,
hence slowing down their processing time in reading the words and thus, hindering and limiting the
reliability of our results..
To fix and improve this for future use, my group should decide as a whole, in what we should all do
consistently and similarly. Like write up a script that all testers in the group would say when reminding the
participants. That way our method of gathering the data would be precise and accurate and thus, our
results and whole investigation would also be accurate and reliable. As a whole team,I though that we have
Pakuranga College Page 10 of 12 2016, AS91265 (v2) Experiments
discussed all the things needed when we were planning and implementing it, but as it turns out, we had
forgotten some of the key controlled variables for this experiment. This is probably because we were more
focused on getting all our 30 data under 1 hour (hence why all of my group tested the participants) which to
be honest, under this time condition, is actually quite limiting as everything is rushed. Thus, also limiting the
accuracy and reliability of our results.
Also, when doing our experiment, my group and I all used our phone’s stopwatch than an actual stopwatch
due to lack of resource. This could’ve limited the reliability in our gathered results as we had different
phone types that might also have a slightly different type of stopwatch function. I didn't realise this until our
experiment was done so I wasn’t able to consider this during my planning period (thus the reason why it
wasn't included in my ‘PLAN’. So in the future, we will use an actual stopwatch for each tester as to ensure
validity and reliability in our results.
Looking back to our experiment and our findings, I was also surprised when I noticed that majority of the
female subjects had a much lower difference in their processing rate when the treatment was introduced in
the reading task. This is shown in my dot box plot of difference where aside from the girl with 28 seconds
difference, all off the data were 12 seconds and below with a range of 8 seconds. Whereas male
participants minimum data was 5 seconds and maximum data was 25 seconds (range 20 seconds). These
findings had made me quite curious whether if gender has played a role in the participants’ processing rate
when cognitive semantic interference is introduced in a reading task in the form of a Stroop Test.
Based on this website, https://www.sbp-journal.com/index.php/sbp/article/view/1477 :“Gender differences
were observed in the color card and color-word card tests but not significantly for the word
card test. The results do substantiate the gender difference in Stroop color and interference.”
With this, in the future, we could investigate this further to see which gender would have a
faster processing rate when a cognitive and semantic interference is introduced in a reading
task via the Stroop Test.
http://www.whatispsychology.biz/about-stroop-effect-definition(©2016 Scientific Journal Publishers Limited. All Rights Reserved.) https://www.sbp-journal.com/index.php/sbp/article/view/1477 (February 24, 2012 by WIP.) https://faculty.washington.edu/chudler/words.html (Copyright © 1996-2015, Eric H. Chudler All Rights Reserved.)
Pakuranga College Page 11 of 12 2016, AS91265 (v2) Experiments
Name: CLASS: JPP Grade: N A M E
NOT ACHIEVED
Multistructural
Thinking (ACHIEVED)
Relational Thinking (MERIT)
Extended Abstract Thinking
(EXCELLENCE)
Marker: ADN N Low High Low High Low High
posing an investigative question about a given experimental situation
✓
planning the experiment by determining appropriate variables and measures
✓
planning the experiment by determining data collection and recording methods
✓
conducting the experiment and collecting data
•
✓
selecting appropriate displays and measures
✓
discussing displays and measures
✓
communicating findings in a conclusion.
✓
Marker’s Judgement
ACHIEVEMENT
Conduct an experiment to investigate a situation using statistical methods involves showing evidence of using each component of the investigation process.
Marker’s Judgement
MERIT
Conduct an experiment to investigate a situation using statistical methods, with justification involves linking components of the process of investigating a situation by experiment to the context, explaining relevant considerations in the investigation process, and supporting findings with statements which refer to evidence gained from the experiment.
Marker’s Judgement
EXCELLENCE
✓
Conduct an experiment to investigate a situation using statistical methods, with statistical insight involves integrating statistical and contextual knowledge throughout the investigation process which may involve reflecting on the process, or considering other relevant variables.
COMMENTS: Impressive – well done! Very good statistical insight shown throughout!
Pakuranga College Page 12 of 12 2016, AS91265 (v2) Experiments