The Alt-Ac Job Beat Newsletter Post 2 10/24/2023

Hi Everyone,

Post 2 of the AltAcademic newsletter! (Again, mea culpa for adding you to this list without permission, you can respond with unsubscribe to this email and I will take you off the list, no hard feelings! I have added many professors I think it would be good to forward to your PhD students as well.)

JOBS

You can bookmark the jobs spreadsheet to check out the specific listings I discuss.

This spreadsheet is not so great to monitor for exact positions for yourself. The goal is moreso to expose people to potential positions, example salary ranges, and things to search for to help you tailor your job search specifically for yourself.

So here I list two different "intel" positions -- if you formerly worked as a crime analyst or in law enforcement, you have good background for these positions. (Knowing you will pass the background is a big plus for these agencies/positions, they have maybe 50%+ attrition at background.)

To find these positions, for the QinetiQ I searched on LinkedIn "OSINT" + Omaha. For NGA I just searched "NGA".

"Analyst" jobs can take on many hats. The UPS role here I think should probably be labelled as "Project Manager", it lists things like Microsoft Office and managing the project backlog (so you won't be crunching numbers in this position). "Project backlog" is agile project management speak for "list of things we need to accomplish". Any academic who has managed students though is well qualified for this position -- don't get fooled by agile buzzwords.

EXAMPLE SCIENTIST

Another federal agency that hires data scientists/analysts is the FBI, check out Karise Curtis (PhD in CJ from SUNY Albany), LinkedIn Profile. Karise got this position shortly before she finished -- don't wait until you have already defended to begin your job search!

TECH ADVICE

Many data oriented jobs have SQL as a skill they want you to have. If you are familiar with merging data sources in stat programs you will be able to pick up SQL fairly quickly. The bigger pain point is that to run SQL code, you need an actual database and to populate it with tables. I have some crime analysis examples in Access, https://apwheele.github.io/Class_CrimeAnalysis/Lab00_ReadMe.html, but note Access has a limited version of SQL (no CTEs, will show an example of those in a future newsletter post).

Here are some code snippets in python to create an in-memory database (using sqlite and pandas). You can then test out SQL code in this environment, to help teach yourself some SQL.

import pandas as pd
import sqlite3

con = sqlite3.connect(":memory:")
tab1 = pd.read_csv('...path to csv...')
tab1.to_sql('table1',index=False,if_exists='replace',con=con)

Then you can read back tables:

sql = 'SELECT * FROM table1'
res = pd.read_sql_query(sql,con)

Or probably of more interest, learning more complicated queries:

sql = '''/* Join example with IN clause */
SELECT
  t1.id
  t1.v1,
  t1.v2,
  t2.v3
FROM table1 AS t1
LEFT JOIN table2 AS t2
  ON t1.id = t2.id
WHERE t2.v4 IN ('A','B','C')
'''

res = pd.read_sql_query(sql,con)

If you just want to execute some SQL statement (that does not return anything), you can just use con.execute('..statement...'):

# demo dropping table
con.execute('DROP TABLE table1;')

Best, Andy Wheeler