Skip to content

Commit

Permalink
Advancment
Browse files Browse the repository at this point in the history
  • Loading branch information
andresnowak committed Dec 20, 2024
1 parent 03a4b5b commit 923f1ae
Show file tree
Hide file tree
Showing 7 changed files with 120 additions and 17 deletions.
132 changes: 115 additions & 17 deletions app/page.js
Original file line number Diff line number Diff line change
Expand Up @@ -416,7 +416,23 @@ export default function Home() {
with a single, comprehensive metric.
</p>
</div>
asoidfjqowiejf
<VariableChooserComponent
Title="Average view count for videos of different sport events"
variables={[
{
datapath:
"data/part_2_plot/avg_view_count_sports_events.csv",
name: "Average view count",
},
]}
>
{(variable) => (
<BarPlotChart
dataPath={variable}
loading={<LoadingSpinner />}
/>
)}
</VariableChooserComponent>
<div className="flex flex-col space-y-2 p-4 w-full max-w-4xl">
<p className="text-lg text-left">
The plot reveals that the <b>FIFA World Cup</b> leads
Expand Down Expand Up @@ -619,9 +635,15 @@ export default function Home() {
First, the following chart visualizes the number of
video uploads for each sport on YouTube over time.
</p>
<h3 className="text-2xl font-bold mb-2">
transition chart will come here
</h3>
</div>
<RacingChartComponent
dataPath={"data/third_plot/sport_transit.csv"}
xName={"Number of videos"}
yName={"Sport category"}
loading={<LoadingSpinner />}
/>
<div></div>
<div className="flex flex-col space-y-2 p-4 w-full max-w-4xl text-lg text-left">
<p>
Interesting observation includes that between 2005 and
2012, hockey dominated YouTube, driven by major events
Expand Down Expand Up @@ -660,15 +682,97 @@ export default function Home() {
</p>
</div>
<div className="flex flex-col space-y-2 p-4 w-full max-w-4xl text-lg text-left">
<h2 className="text-2xl font-bold mb-2">
Difference in difference Next
</h2>
<h3 className="text-1xl font-bold mb-2">
Difference in difference
</h3>
<p>
We explore the how much sport-events affect the
engagement
We investigated how specific sports events contribute to
engagement on YouTube using Difference-in-Differences
(DID). The analysis focused on data from 2017-2018, a
period when YouTube was becoming more mature. We
examined how the delta_view (change in views) of videos
in the period before and after sports events changed for
five sports: football, skiing, basketball, hockey, and
rugby. The corresponding events were: FIFA 2018, Winter
Olympics 2018, NBA Finals 2018, Stanley Cup Finals 2018,
and Rugby World Cup Sevens 2018. The control group was
the average delta_view of videos from sports unrelated
to the events, and the sport group was the average
delta_view of videos related to the event in question.
</p>
<p>
The key finding is that not all sports events contribute
to increased engagement in their respective sports.
</p>
<p>
<strong>Football and Skiing:</strong>
These two sports show the most significant impact from
their events, as the graphs clearly demonstrate.
However, when examining the p-values, football shows all
p-values below 0.05, indicating a strong, positive
impact from the FIFA 2018 event, with statistically
significant coefficients for both the treatment and
interaction terms. On the other hand, skiing shows a
weaker and less consistent pattern, with the Winter
Olympics having little effect on engagement and no
significant interaction term. It&#39;s important to note
that for this DID analysis, we examined the data for 4
months before and 4 months after the events. Therefore,
we can conclude that user engagement for skiing did not
sustain for long after the Winter Olympics.
</p>
<p>
<strong>Basketball, Hockey, and Rugby:</strong>
For these sports, the events did not show a substantial
increase in engagement. Particularly surprising was the
lack of change in engagement after the NBA Finals.
Despite the large number of basketball-related videos,
there was no noticeable shift in engagement. This could
be due to the NBA Finals being a U.S.-centric event, not
a global competition, and since it is a culmination of
the season, engagement may not increase significantly
afterward. For hockey, the post-event engagement
actually decreased, which is puzzling. This could be
because the Stanley Cup Finals took place in June, a
time when interest in winter sports, like hockey,
typically wanes, suggesting that the timing of the event
is critical in sustaining user interest.
</p>
<p>
From these findings, it is clear that sports events do
not always affect engagement on YouTube. However, for
sports like football and skiing, the effect was
noticeable. Additionally, it can be concluded that for
minor-sports to capture more attention, the event needs
to be held on a global scale and during an appropriate
season for the sport.
</p>
</div>
images
<VariableChooserComponentImage
Title=""
variables={[
{
datapath: "data/diff_in_diff/basketball_DID.png",
name: "Basketball DID",
},
{
datapath: "data/diff_in_diff/football_DID.png",
name: "Football DID",
},
{
datapath: "data/diff_in_diff/hockey_DID.png",
name: "Hockey DID",
},
{
datapath: "data/diff_in_diff/rugby_DID.webp",
name: "Rugby DID",
},
{
datapath: "data/diff_in_diff/skiing_DID.png",
name: "Skiing DID",
},
]}
/>
<div className="flex flex-col space-y-2 p-4 w-full max-w-4xl text-lg text-left">
<h1 className="text-3xl font-bold mb-2">
<strong>Part 4: User Engagement Over Time</strong>
Expand Down Expand Up @@ -827,12 +931,6 @@ export default function Home() {
title="Create Next App"
text="Lorem ipsum dolor sit, amet consectetur adipisicing elit. Ipsam autem quibusdam, delectus in corrupti, ab impedit magni iure eveniet aliquid soluta neque quisquam ducimus dolores ex suscipit pariatur. Voluptatibus, exercitationem? Lorem ipsum, dolor sit amet consectetur adipisicing elit. Perspiciatis repellat iure tenetur similique nemo soluta velit voluptate."
/>
<RacingChartComponent
dataPath={"data/third_plot/sport_transit.csv"}
xName={"Number of videos"}
yName={"Sport category"}
loading={<LoadingSpinner />}
/>
<SubTitleText
title="Create Next App"
text="Lorem ipsum dolor sit, amet consectetur adipisicing elit. Ipsam autem quibusdam, delectus in corrupti, ab impedit magni iure eveniet aliquid soluta neque quisquam ducimus dolores ex suscipit pariatur. Voluptatibus, exercitationem? Lorem ipsum, dolor sit amet consectetur adipisicing elit. Perspiciatis repellat iure tenetur similique nemo soluta velit voluptate."
Expand All @@ -852,7 +950,7 @@ export default function Home() {
]}
>
{(variable) => (
<BoxPlotChart
<BarPlotChart
datapath={variable}
loading={<LoadingSpinner />}
/>
Expand Down
Binary file added public/data/diff_in_diff/basketball_DID.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added public/data/diff_in_diff/football_DID.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added public/data/diff_in_diff/hockey_DID.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added public/data/diff_in_diff/rugby_DID.webp
Binary file not shown.
Binary file added public/data/diff_in_diff/skiing_DID.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
5 changes: 5 additions & 0 deletions public/data/part_2_plot/avg_view_count_sports_events.csv
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
Sport event,Average view count
FIFA World Cup,426137.01531862747
NBA Finals,131716.1448692153
Olympics,94457.54289503214
Others,59817.37653166949

0 comments on commit 923f1ae

Please sign in to comment.