Have a personal or library account? Click to login
Survey vs Scraped Data: Comparing Time Series Properties of Web and Survey Vacancy Data Cover

Survey vs Scraped Data: Comparing Time Series Properties of Web and Survey Vacancy Data

IZA Journal of Labor Economics
Volume 8 (2019): Issue 1 (June 2019)
By: Pablo de Pedraza,  Stefano Visintin,  Kea Tijdens and  Gábor Kismihók  
Open Access
|Sep 2019

Figures & Tables

Figure 1

NSO and web-extracted data on vacancies in the Dutch economy, 2007–2014.NSO, National Statistics Office

Figure 2

Web and NSO quarterly vacancies, time series decomposition: TC components.NSO, National Statistics Office; TC, trend–cycle

Figure 3

Web and NSO quarterly vacancies, time series decomposition: seasonal. NSO, National Statistics Office

Figure 4

NSO and web-extracted, differentiated data on vacancies in the Dutch economy, 2007–2014. NSO, National Statistics Office

Figure 5

Autocorrelation of the original and differentiated series, top and bottom rows, respectively, lags expressed in years

Figure 6

Cross-correlation between NSO and web-extracted data on vacancies in the Dutch economy. Raw data (top) and differentiated data (bottom). NSO, National Statistics Office

Figure 7

Cross-spectral density, original series (top), and differentiated series (bottom)

Figure 8

Squared coherency, original series (top), and differentiated series (bottom)

Unit root test statistics

Unit root test statistics
ADF testADF test (modified)PP testZA test (and critical values)
Web-4.6675*** (4)-1.0229 (3)-17.177* (2)-6.1109 ***(6)0.01= -5.34 0.05= -4.8
0.1= -4.58
NSO-2.884 (4)-1.6402 (3)-8.0429 (2)-5.2097**(8)0.01= -5.34 0.05= -4.8
0.1= -4.58

Web and NSO quarterly vacancies, time series decomposition: TC components

TimeWeb trendNSO trend% Difference (decimal form)
2006Q41993442450270.19
2007Q11937592488980.22
2007Q21919092511240.24
2007Q31914482500380.23
2007Q41947382471370.21
2008Q11893342448790.23
2008Q21668672384820.30
2008Q31374992222690.38
2008Q41260551991270.37
2009Q11269301771550.28
2009Q21254081655670.24
2009Q31171551624850.28
2009Q41081891616010.33
2010Q11038091619660.36
2010Q21052351649120.36
2010Q31091671698500.36
2010Q41156621743780.34
2011Q11250331762050.29
2011Q21330921759010.24
2011Q31331491723500.23
2011Q41247281650960.24
2012Q11165271571010.26
2012Q21146961515300.24
2012Q31136121474650.23
2012Q41110371426520.22
2013Q11113031387760.20
2013Q21142181385190.18
2013Q31170961419020.17
2013Q41196651467360.18
2014Q11285751516750.15
2014Q21382351571640.12
2014Q31495381627320.08
DOI: https://doi.org/10.2478/izajole-2019-0004 | Journal eISSN: 2193-8997
Journal RSS Feed
Language: English
Published on: Sep 13, 2019
Published by: Sciendo
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year
Keywords:
web crawling,
statistical inference,
time series,
vacancies,
Labor demand,
data collection
Related subjects:
Business and economics,
Political economics,
Microeconomics,
Macroeconomics,
Economic policy,
Mathematics and statistics for economists,
Econometrics

© 2019 Pablo de Pedraza, Stefano Visintin, Kea Tijdens, Gábor Kismihók, published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 License.

Previous articleVolume 8 (2019): Issue 1 (June 2019)Next article