Analyzing data from POD's first survey to the iSchool¶
The polling and open data initiative at UW (POD) sent out its first poll to students pursuing a degree in informatics, receiving 42 replies. I analyzed some of the data regarding the different tracks under the iSchool, plans after college, and grad school interest.
Some of the data were visualized using matplotlib to find trends, while visuals to be published at POD used Flourish.
Cleaning the data¶
In [398]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
In [399]:
poll = pd.read_excel("poll1.xlsx")
poll.head(8)
Out[399]:
In [437]:
table = poll[7:]
table.columns = poll.iloc[6]
table.loc[:, "total"] = np.ones(table.shape[0])
table.loc[:, "researchInt"] = 0
table.loc[table.research == "Yes", "researchInt"] = 1
table.head()
Out[437]:
In [424]:
# Adding seperate categorical variables for each option under "plans"
table.loc[:, "workInTech"] = table.plans.str.contains("Work in information technology").astype(int)
table.loc[:, "nonprofit"] = table.plans.str.contains("Work in a nonprofit organization").astype(int)
table.loc[:, "travel"] = table.plans.str.contains("Travel abroad").astype(int)
table.loc[:, "educate"] = table.plans.str.contains("Work in education").astype(int)
table.loc[:, "unrelated"] = table.plans.str.contains("Work in a field unrelated to informatics").astype(int)
table.loc[:, "masters"] = table.plans.str.contains("Pursue a Master").astype(int)
table.loc[:, "phd"] = table.plans.str.contains("Pursue a PhD").astype(int)
In [ ]:
Graduation Year vs GradSchool Interest¶
In [402]:
table.masters.sum()
Out[402]:
In [403]:
table.phd.sum()
Out[403]:
In [404]:
# Nobody signified phd but not masters; so we can use "msaters" as interest in Grad School
table.loc[(table.phd == 1) & (table.masters == 0)]
Out[404]:
In [405]:
mastersInterest = table.groupby("GradYear").masters.sum()
_ = plt.bar(mastersInterest.index.astype(str), mastersInterest)
In [406]:
table.groupby("GradYear").researchInt.sum()
Out[406]:
In [407]:
table.loc[table.track == "Custom"].groupby("GradYear").total.sum()
Out[407]:
As expected, upperclassmen have more research experience, and are more interested in grad school than underclassmen.
Research Areas¶
In [408]:
researchers = table.loc[table.researchInt == 1]
In [409]:
researchAreas = researchers.groupby(researchers.researchField).total.sum()
pieLabels = researchAreas.index
_ = plt.pie(researchAreas, labels=pieLabels)
In [410]:
print(researchers.shape[0], "total students with research experience")
Custom tracked students¶
In [411]:
customTracked = researchers.loc[researchers.track == "Custom"]
researchAreasC = customTracked.groupby(customTracked.researchField).total.sum()
pieLabelsC = researchAreasC.index
_ = plt.pie(researchAreasC, labels=pieLabelsC)
In [412]:
researchAreasC
Out[412]:
research vs grad school plans¶
In [422]:
researchVGrad = table.groupby("researchInt").sum()[["total", "masters"]]
researchVGrad
Out[422]:
Correlation between interset in pursuing grad school and research experience seem to align, without neccessarily implying causation one way or the other.
Plans after the iSchool¶
In [426]:
planSums = table[["workInTech", "nonprofit", "travel", "educate", "unrelated", "masters", "phd", "researchInt"]].sum()
planSums
Out[426]:
In [427]:
trackSums = table.groupby("track").total.sum()
trackSums
Out[427]:
In [433]:
summaryTable = table.groupby("track").sum().T
summaryTable["sums"] = planSums
summaryTable
Out[433]:
In [434]:
tracks = summaryTable.T
tracks["sums"] = trackSums
tracks
Out[434]:
In [436]:
table.to_csv("table.csv")
summaryTable.to_csv("trackPlansSummaryTable.csv")
tracks.to_csv("tracksSummary.csv")