Careers Step 1, Wave 2: Fraud dection & cleaning

Basic count Wave 2 completes

Qualtrics good completes : 4041

WPA good completes : 3815

• Age: within age range (21 <= age <= 66)
  
• Education: min Bachelors degree

---

---

---

Geo-locatoin

Lat-Long: ex-US

ex-US response:
- 5 Nigreria

US map

---

Declared state of residence and lat-long position

Response farms will many repeated lat-long positions with varrying declared state of residence

---

Repeat Lat-Long.

• 21 response from the same Lat-Long in rural Kansas (free text is short & similar) in WAVE 1

---

---

Dream Job:

DreamJob	Freq	perc
No	1588	39.3
Yes	2453	60.7

Do Over:

Do_over	Freq	perc
No	1649	40.81
Yes	2392	59.19

---

---

Quality Checks - Step 2: Wharton Review

Our review will follow a few steps: First, we will check for any suspicious latitude-longitude coordinates in the US. We will flag two types of responses—excessive repeats from the same lat-long and/or lat-long coordinates that don’t match the state of residence indicated in the survey. Any responses flagged as suspicious will be checked for other signs of suspicious responding (see below). This will catch cases of repeat survey-takers, like in the pilot data when there were 21 responses from the same lat-long in rural Kansas, all with similar short free-text responses.
Next, we will do a and check for a few improbable / inconsistent types of responding. Any responses with 2 or more of these flags will be rejected:

• If age minus years of post-college work experience is < 18.
• If total years of work experience minus tenure in current job is negative.
  
• If age at year of completing Bachelor’s degree is < 20.
• If age at year of completing doctoral degree is < 25.
• Manual review of job titles (flagging any highly unusual text)

---

Fix / clean variables used in QC checks

work_exp_fix

work_exp: How many years of post-college work experience do you have?

mutate(work_exp_fix = case_when(is.na(work_exp) ~ NA,
                                work_exp == "Less than 1" ~ 0.5,
                                as.numeric(work_exp) < (66-17) ~ as.numeric(work_exp),
                                as.numeric(work_exp) >= (66-17) ~ NA,
                               .default = NA  ) ) 
                              

job_tenure_fix

job_tenure: For how many years have you been employed by your current employer? Please provide the total number of years since you first started working for this employer, even if you have been promoted or changed jobs with this same employer since then.

       mutate(job_tenure_fix = case_when(job_tenure == "Less than 1" ~ 0.5,
                                 as.numeric(job_tenure) < (66-17) ~ as.numeric(job_tenure),
                                 .default = NA  ) ) 

graduate_fix

graduate: In what year did you complete your Bachelor’s degree?

     mutate(graduate_fix = case_when(graduate== "Before 1960" ~ NA, # too old
                              as.numeric(graduate) < (2024-66+15) ~ NA, # too old
                              as.numeric(graduate) >= (2024-66+15) ~ as.numeric(graduate) )) 

graduate_begin_fix

college begin (year): In what year did you enter college for your Bachelor’s degree?

mutate(college_begin_fix = case_when(college_begin== "Before 1960" ~ NA, # too old
                                   as.numeric(college_begin) < (2024-66+15) ~ NA, # too old
                                   as.numeric(college_begin) >= (2024-66+15) ~ as.numeric(college_begin) )) %>%

---

---

Variable pair conflicts:

If age minus years of post-college work experience is < 18.

C1: age_start_work: age minus years of post-college work experience is < 18 (age-work_exp_fix)

     work_exp_start_calc = age - work_exp_fix, # age that you start working post college
     

work_exp_start_calc	Freq
0	4010
1	31

---

If total years of work experience minus tenure in current job is negative.

C2: total years of work experience minus tenure in current job is negative

              work_exp_L_job_tenure = work_exp_fix - job_tenure_fix,

Work_exper_L_tenure	Freq
0	2690
1	1351

---

If age at year of completing Bachelor’s degree is < 20.

This is cover by the hard age cut. age >= 21

---

If age at year of completing doctoral degree is < 25.

C4: If age at year of completing doctoral degree is < 25.

     BAD_Educ_Age = case_when(education== "Professional degree (e.g., MD, DDS, DVM, JD)" & age <= 25 ~ BAD_num,    
                               education== "Doctorate degree (e.g., PhD, EdD)" & age <= 25 ~ BAD_num, 
                               .default = GOOD_num),

Manual review of job titles (flagging any highly unusual text)

---

Other possible metrics:

Age at college graduation {age- (2024-graduate)}

---

Duration

---

Job & Employers

under the line is improbable / impossible?

---

Repeate Lat - Long (red lines are starting points of flag)

---

Per state, respsonse beyond expected base on US census adult populations.

---

---

BOTs

bots_num	job_title_lower	state	gender	age
1	it software	MD	Man	39
2	techonology development software	NM	Man	41
3	i don’t have a job title since i’m an ai created by openai to assist with a wide range of queries and tasks. my purpose is to help users by providing information, answering questions, and offering support in various topics.	CA	Man	37
4	computer software	VI	Man	43
5	ai language model, conversational assistant.	AZ	Man	61

bots_num	biggest_influence	state	gender	age
1	researchers and engineers at openAI	MD	Man	39
2	My initial “career choice,” if you could call it that, was shaped by a combination of factors. The development of AI models like me was influenced by a drive to understand and replicate human language and thought processes. My “training” involved learning from vast amounts of text data, so I guess you could say that the diversity and depth of human knowledge, as well as the technological advancements in machine learning, were significant influences. It’s less about a personal choice and more about the evolution of technology and the goals set by researchers and developers.	NM	Man	41
3	, I don’t have a personal career or experiences to draw from. However, for many people, the biggest influences on their initial career choice can include factors like personal interests, skills, family expectations, educational background, or significant life experiences. If you’re reflecting on your own career path, these might be things to consider.	CA	Man	37
4	meaningful ways. The driving forces behind my development were the desire to improve human-computer interaction and make information more accessible and useful. What about you? What factors influenced your career choice	VI	Man	43
5	Passion for the field inspired me.	AZ	Man	61

---

bots_num	state	age	gender	dreamjob	career_sat_1	income_house	personal_income	marital	education	graduate	BAD_repeate_Lat_Long	BAD_dist_to_state	repeat_LL_freq	Duration (in seconds)
1	MD	39	Man	No	Agree Strongly	$100,000 - $149,999	$100,000 - $149,999	Married or domestic partnership	Bachelor’s degree (e.g., BA, BS)	2008	0.0	0	3	677
2	NM	41	Man	No	Agree	$100,000 - $149,999	$100,000 - $149,999	Married or domestic partnership	Master’s degree (e.g., MA, MS, MEd)	2008	0.7	1	5	1179
3	CA	37	Man	Yes	Agree Strongly	$100,000 - $149,999	$100,000 - $149,999	Married or domestic partnership	Master’s degree (e.g., MA, MS, MEd)	2007	1.0	1	15	1530
4	VI	43	Man	Yes	Agree Strongly	$150,000 - $199,999	$150,000 - $199,999	NA	Master’s degree (e.g., MA, MS, MEd)	2010	0.0	1	1	960
5	AZ	61	Man	Yes	Agree Strongly	$100,000 - $149,999	$100,000 - $149,999	Married or domestic partnership	Master’s degree (e.g., MA, MS, MEd)	2004	0.7	1	12	925

---

bots_num	state	age
1	MD	39
2	NM	41
3	CA	37
4	VI	43
5	AZ	61

---

---