SlideShare a Scribd company logo
1 of 29
Danny Pająk   Research Factory SoSe 2011 THE SHADOW OF HIERARCHY How to Sample a Hidden Population of Former Employees? 7. Juni 2011 Fotolia.com: Cartoon Crowd, Complex System © higyou #15575655 Fotolia.com:  Kommunikation, Business und Logistik © ag visuell #16575699
Outline ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29
Narrowing down the research topic ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29
The shadow of hierarchy Victims node link or tie ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29 Survivors White-collar male Blue-collar female
Institute for Management & Marketing (M&M) Research Factory SoSe 2011 7. Juni 2011 -  Danny Pająk Folie  /29
Folie  /29 Institute for Management & Marketing (M&M) Research Factory SoSe 2011 7. Juni 2011 -  Danny Pająk
http://www.ima-research.eu/projekt/verlagerung/datenbank/ Folie  /29
Folie  /29
Studying hidden populations ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Source: Kendall, 2006, p. 5 Folie  /29
Respondent-driven sampling (RDS) ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29
Snowball sampling biases and RDS “solutions” Respondents may refer to an unlimited number of peers Social network properties are ignored Respondents refer ,  surveyors must find referred Convenience sample – analysis limited to proportions of sample ,  not generalisable ,[object Object],[object Object],[object Object],[object Object],Only members accessible to “outsider” participate Limiting recruitment coupons to individuals limits clusters ,  thereby reducing recruitment bias and high homophily ,[object Object],[object Object],Peers recruit peers ,  which includes ability to exert social influence where surveyors likely have none; surveyors remain in office Probability of selection is unknown Collect size of peer network to calculate probability of selection within network ,  i.e. sampling weight ,  use known network properties to account for clustering effects Bias Issue Respondent Driven Sampling “Solution” Snowball Sampling Source: Johnston   & Sabin, 2010, p. 39 Folie  /29
Traditional sampling and estimation and respondent-driven sampling Source: Salganik   & Heckathron ,  2004, p. 200 ,[object Object],[object Object],[object Object],[object Object],Folie  /29 Population Sample Collection Estimation Population Sample Social Network Estimation Estimation Collection
How to decide whether RDS is a suitable sampling method? Folie  /29 Based on: Johnston   & Sabin, 2010, p. 45
Advantages of RDS ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29 Source: Wang, 2004; Wang et al., 2005
Limitations  of RDS ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29 Source: Wang, 2004; Wang et al., 2005
Feasibility study Folie  /29 Manufacturer of electrical equipment which relocated its entire workforce from Berlin to a rural area in the federal state of Lower Saxony in 2006
Giant component Folie  /29 ,[object Object],[object Object],[object Object],[object Object]
Degree distribution 1/2 Folie  /29 ,[object Object],[object Object]
Degree distribution 2/2 ,[object Object],[object Object],[object Object],Folie  /29
Memory gaps or masking Folie  /29
Outdegree layout Folie  /29
Exemplary RDS chains Folie  /29 in-degree/out-degree 15/15 10/10 6/4 4/12 6/3 11/11 4/.. 2/.. 1/.. 2/.. 2/.. 3/.. 4/.. 1/.. 10/7 2/.. 1/.. 1/1 1/.. 6/6 4/.. 16/15 11/10 8/8 7/6 2/.. 5/5 2/.. 7/.. 20/19 seeds wave 1 wave 2 wave 3 wave 4 wave 5 wave 6 oversampling undersampling wave n female male blue- collar withe- collar victims sur- vivors German Turkish others 0 4 50.0% 50.0% 50.0% 50.0% 100.0% 0.0% 100.0% 0.0% 0.0% 1 10 50.0% 50.0% 30.0% 70.0% 80.0% 20.0% 100.0% 0.0% 0.0% 2 18 61.1% 38.9% 44.4% 55.6% 83.3% 16.7% 94.4% 5.6% 0.0% 3 21 57.1% 42.9% 38.1% 61.9% 81.0% 19.0% 90.5% 4.8% 4.8% 4 23 52.2% 47.8% 34.8% 65.2% 82.6% 17.4% 91.3% 4.3% 4.3% 5 24 50.0% 50.0% 37.5% 62.5% 83.3% 16.7% 91.7% 4.2% 4.2% 6 26 46.2% 53.8% 34.6% 65.4% 84.6% 15.4% 92.3% 3.8% 3.8% True Population 90 58.9% +12.7% 41.1% -12.7% 57.8% +23.2% 42.2% -23.2% 76.7% -7.9% 23.3% +7.9% 64.4% -27.9% 25.6% +21.7% 10.0% +6.2% Giant Component 75 52.0% +5.8% 48.0% -5.8% 50.7% +16.1% 49.3% -16.1% 72.0% -12.6% 28.0% +12.6% 72.0% -20.3% 22.7% +18.9% 5.3% +1.5%
Homophily ,[object Object],Folie  /29 Male (41.1%) Female (58.9%) 37.5% 40.0% 62.5% 60.0% Blue-collar workers (57.8) White-collar workers (42.2%) 76.9% 53.8% 23.1% 46.2%
Exemplary recruiting behavior by citizenship German (64.4%) Turkish (25.6%) 8% 20% Yugoslav (5.6%) Others (4.4%) 5% 5% 4% 2% 60% 2% 50% 1% 10% 10% Folie  /29 88% 35% 75% 25%
Recruiting pattern during different recruitment waves  Starting point: All German subjects True German Population 64.4% True Turkish Population 25.6% Folie  /29
Theoretical growth of recruitment Folie  /29 Source: Wang et al., 2005, p. 152
Using XING to find the initial seeds for RDS ,[object Object],[object Object],[object Object],[object Object],Folie  /24 Source: Xing (http://www.xing.com) - Modified and anonymized profile screenshot, retrieved May 9, 2011.
References ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],Folie  /29
Danny Pająk  - pajak@europa-uni.de Lehrstuhl für Internationales Management Tel.: +49-(0)335-5534-2813 Europa-Universität Viadrina Große Scharrnstraße 59 15230 Frankfurt (Oder) Folie  /29 Thank you

More Related Content

Similar to THE SHADOW OF HIERARCHY - HOW TO SAMPLE A HIDDEN POPULATION OF FORMER EMPLOYEES?

Social network analysis & Big Data - Telecommunications and more
Social network analysis & Big Data - Telecommunications and moreSocial network analysis & Big Data - Telecommunications and more
Social network analysis & Big Data - Telecommunications and more
Wael Elrifai
 
On Machine Learning and Data Mining
On Machine Learning and Data MiningOn Machine Learning and Data Mining
On Machine Learning and Data Mining
butest
 

Similar to THE SHADOW OF HIERARCHY - HOW TO SAMPLE A HIDDEN POPULATION OF FORMER EMPLOYEES? (20)

Negotiated Studies Presentation on Social Network Analysis of Knowledge Networks
Negotiated Studies Presentation on Social Network Analysis of Knowledge NetworksNegotiated Studies Presentation on Social Network Analysis of Knowledge Networks
Negotiated Studies Presentation on Social Network Analysis of Knowledge Networks
 
Social Νetworks Data Mining
Social Νetworks Data MiningSocial Νetworks Data Mining
Social Νetworks Data Mining
 
Everything you always wanted to know about Synthetic Data
Everything you always wanted to know about Synthetic DataEverything you always wanted to know about Synthetic Data
Everything you always wanted to know about Synthetic Data
 
Discovering Influential User by Coupling Multiplex Heterogeneous OSN’S
Discovering Influential User by Coupling Multiplex Heterogeneous OSN’SDiscovering Influential User by Coupling Multiplex Heterogeneous OSN’S
Discovering Influential User by Coupling Multiplex Heterogeneous OSN’S
 
Big Data Analytics : A Social Network Approach
Big Data Analytics : A Social Network ApproachBig Data Analytics : A Social Network Approach
Big Data Analytics : A Social Network Approach
 
02 Network Data Collection
02 Network Data Collection02 Network Data Collection
02 Network Data Collection
 
02 Network Data Collection (2016)
02 Network Data Collection (2016)02 Network Data Collection (2016)
02 Network Data Collection (2016)
 
Analysing a Complex Agent-Based Model Using Data-Mining Techniques
Analysing a Complex Agent-Based Model  Using Data-Mining TechniquesAnalysing a Complex Agent-Based Model  Using Data-Mining Techniques
Analysing a Complex Agent-Based Model Using Data-Mining Techniques
 
WSDM 2018 Tutorial on Influence Maximization in Online Social Networks
WSDM 2018 Tutorial on Influence Maximization in Online Social NetworksWSDM 2018 Tutorial on Influence Maximization in Online Social Networks
WSDM 2018 Tutorial on Influence Maximization in Online Social Networks
 
GLAM Survey presentation OpenSym/WikiSym 2013
GLAM Survey presentation OpenSym/WikiSym 2013GLAM Survey presentation OpenSym/WikiSym 2013
GLAM Survey presentation OpenSym/WikiSym 2013
 
Social network analysis & Big Data - Telecommunications and more
Social network analysis & Big Data - Telecommunications and moreSocial network analysis & Big Data - Telecommunications and more
Social network analysis & Big Data - Telecommunications and more
 
Social network analysis (SNA) - Big data and social data - Telecommunications...
Social network analysis (SNA) - Big data and social data - Telecommunications...Social network analysis (SNA) - Big data and social data - Telecommunications...
Social network analysis (SNA) - Big data and social data - Telecommunications...
 
2007.02500.pdf
2007.02500.pdf2007.02500.pdf
2007.02500.pdf
 
Fundamentals of Deep Recommender Systems
 Fundamentals of Deep Recommender Systems Fundamentals of Deep Recommender Systems
Fundamentals of Deep Recommender Systems
 
Robust Expert Finding in Web-Based Community Information Systems
Robust Expert Finding in Web-Based Community Information SystemsRobust Expert Finding in Web-Based Community Information Systems
Robust Expert Finding in Web-Based Community Information Systems
 
Introduction of abm
Introduction of abmIntroduction of abm
Introduction of abm
 
Social Network Analysis
Social Network AnalysisSocial Network Analysis
Social Network Analysis
 
On Machine Learning and Data Mining
On Machine Learning and Data MiningOn Machine Learning and Data Mining
On Machine Learning and Data Mining
 
Social Network Analysis Helps Telecommunication Firms
Social Network Analysis Helps Telecommunication FirmsSocial Network Analysis Helps Telecommunication Firms
Social Network Analysis Helps Telecommunication Firms
 
Detection and Minimization Influence of Rumor in Social Network
Detection and Minimization Influence of Rumor in Social NetworkDetection and Minimization Influence of Rumor in Social Network
Detection and Minimization Influence of Rumor in Social Network
 

Recently uploaded

Salient Features of India constitution especially power and functions
Salient Features of India constitution especially power and functionsSalient Features of India constitution especially power and functions
Salient Features of India constitution especially power and functions
KarakKing
 
The basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptxThe basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptx
heathfieldcps1
 

Recently uploaded (20)

Interdisciplinary_Insights_Data_Collection_Methods.pptx
Interdisciplinary_Insights_Data_Collection_Methods.pptxInterdisciplinary_Insights_Data_Collection_Methods.pptx
Interdisciplinary_Insights_Data_Collection_Methods.pptx
 
Micro-Scholarship, What it is, How can it help me.pdf
Micro-Scholarship, What it is, How can it help me.pdfMicro-Scholarship, What it is, How can it help me.pdf
Micro-Scholarship, What it is, How can it help me.pdf
 
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
 
Plant propagation: Sexual and Asexual propapagation.pptx
Plant propagation: Sexual and Asexual propapagation.pptxPlant propagation: Sexual and Asexual propapagation.pptx
Plant propagation: Sexual and Asexual propapagation.pptx
 
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
 
Salient Features of India constitution especially power and functions
Salient Features of India constitution especially power and functionsSalient Features of India constitution especially power and functions
Salient Features of India constitution especially power and functions
 
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptxHMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
HMCS Vancouver Pre-Deployment Brief - May 2024 (Web Version).pptx
 
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptxBasic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
 
The basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptxThe basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptx
 
Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)
 
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptxOn_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
 
Single or Multiple melodic lines structure
Single or Multiple melodic lines structureSingle or Multiple melodic lines structure
Single or Multiple melodic lines structure
 
How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdf
 
How to setup Pycharm environment for Odoo 17.pptx
How to setup Pycharm environment for Odoo 17.pptxHow to setup Pycharm environment for Odoo 17.pptx
How to setup Pycharm environment for Odoo 17.pptx
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptx
 
How to Add New Custom Addons Path in Odoo 17
How to Add New Custom Addons Path in Odoo 17How to Add New Custom Addons Path in Odoo 17
How to Add New Custom Addons Path in Odoo 17
 
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptxHMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and Modifications
 
General Principles of Intellectual Property: Concepts of Intellectual Proper...
General Principles of Intellectual Property: Concepts of Intellectual  Proper...General Principles of Intellectual Property: Concepts of Intellectual  Proper...
General Principles of Intellectual Property: Concepts of Intellectual Proper...
 

THE SHADOW OF HIERARCHY - HOW TO SAMPLE A HIDDEN POPULATION OF FORMER EMPLOYEES?

  • 1. Danny Pająk Research Factory SoSe 2011 THE SHADOW OF HIERARCHY How to Sample a Hidden Population of Former Employees? 7. Juni 2011 Fotolia.com: Cartoon Crowd, Complex System © higyou #15575655 Fotolia.com: Kommunikation, Business und Logistik © ag visuell #16575699
  • 2.
  • 3.
  • 4.
  • 5. Institute for Management & Marketing (M&M) Research Factory SoSe 2011 7. Juni 2011 - Danny Pająk Folie /29
  • 6. Folie /29 Institute for Management & Marketing (M&M) Research Factory SoSe 2011 7. Juni 2011 - Danny Pająk
  • 9.
  • 10.
  • 11.
  • 12.
  • 13. How to decide whether RDS is a suitable sampling method? Folie /29 Based on: Johnston & Sabin, 2010, p. 45
  • 14.
  • 15.
  • 16. Feasibility study Folie /29 Manufacturer of electrical equipment which relocated its entire workforce from Berlin to a rural area in the federal state of Lower Saxony in 2006
  • 17.
  • 18.
  • 19.
  • 20. Memory gaps or masking Folie /29
  • 22. Exemplary RDS chains Folie /29 in-degree/out-degree 15/15 10/10 6/4 4/12 6/3 11/11 4/.. 2/.. 1/.. 2/.. 2/.. 3/.. 4/.. 1/.. 10/7 2/.. 1/.. 1/1 1/.. 6/6 4/.. 16/15 11/10 8/8 7/6 2/.. 5/5 2/.. 7/.. 20/19 seeds wave 1 wave 2 wave 3 wave 4 wave 5 wave 6 oversampling undersampling wave n female male blue- collar withe- collar victims sur- vivors German Turkish others 0 4 50.0% 50.0% 50.0% 50.0% 100.0% 0.0% 100.0% 0.0% 0.0% 1 10 50.0% 50.0% 30.0% 70.0% 80.0% 20.0% 100.0% 0.0% 0.0% 2 18 61.1% 38.9% 44.4% 55.6% 83.3% 16.7% 94.4% 5.6% 0.0% 3 21 57.1% 42.9% 38.1% 61.9% 81.0% 19.0% 90.5% 4.8% 4.8% 4 23 52.2% 47.8% 34.8% 65.2% 82.6% 17.4% 91.3% 4.3% 4.3% 5 24 50.0% 50.0% 37.5% 62.5% 83.3% 16.7% 91.7% 4.2% 4.2% 6 26 46.2% 53.8% 34.6% 65.4% 84.6% 15.4% 92.3% 3.8% 3.8% True Population 90 58.9% +12.7% 41.1% -12.7% 57.8% +23.2% 42.2% -23.2% 76.7% -7.9% 23.3% +7.9% 64.4% -27.9% 25.6% +21.7% 10.0% +6.2% Giant Component 75 52.0% +5.8% 48.0% -5.8% 50.7% +16.1% 49.3% -16.1% 72.0% -12.6% 28.0% +12.6% 72.0% -20.3% 22.7% +18.9% 5.3% +1.5%
  • 23.
  • 24. Exemplary recruiting behavior by citizenship German (64.4%) Turkish (25.6%) 8% 20% Yugoslav (5.6%) Others (4.4%) 5% 5% 4% 2% 60% 2% 50% 1% 10% 10% Folie /29 88% 35% 75% 25%
  • 25. Recruiting pattern during different recruitment waves Starting point: All German subjects True German Population 64.4% True Turkish Population 25.6% Folie /29
  • 26. Theoretical growth of recruitment Folie /29 Source: Wang et al., 2005, p. 152
  • 27.
  • 28.
  • 29. Danny Pająk - pajak@europa-uni.de Lehrstuhl für Internationales Management Tel.: +49-(0)335-5534-2813 Europa-Universität Viadrina Große Scharrnstraße 59 15230 Frankfurt (Oder) Folie /29 Thank you

Editor's Notes

  1. Could gain retribution for the organization’s actions through lawsuits. media campaigns. or sabotage. In general. justice concerns discussed previously appear to be important factors for those who lose their jobs during downsizing. although the form of these justice effects is different than among employees who remain following downsizing. Interviews with displaced workers from a number of industries demonstrated that lack of advanced notification and insufficient severance pay were common complaints (Leana & Feldman. 1992). or equivalently. downsized workers report adequate severance was a major reason they viewed their former employers positively (Goldenberg & Kline. 1997). Auf jede 3-4 Verlagerung folgt eine Rückverlagerung. Häufig wird dabei auf ehemalige Mitarbeiter zurückgegriffen. Bei Verlagerungen fallen die Jobs im Gegensatz zum Downsizing nicht weg Gibt es Unterschiede bezüglich der Zielregion der Verlagerung sowie dem Grund der Verlagerung? In Downsizing literature individuals may accept the explanations and justifications for lay-offs as legitimate . and thereby reason that the organization is reacting in a fair and appropriate manner given the surrounding conditions and circumstances that may be beyond the control of the organization (e.g. poor economic conditions).
  2. Institutional Sampling (Gewerkschaften. Outplacementberatung. ...) Time/ Space Sampling (Firma zum Zeitpunkt der Verlagerung) Targeted Sampling (Arbeitsämter) Chain-Referral Sampling (Convenience Sample – keine Rückschlüsse auf Grundgesamtheit)
  3. Methode unterstellt. dass Mitglieder der zu untersuchenden Gruppe in sozialen Netzwerken verbunden sind. Implication: Making chain-referral sampling a form of probability sampling requires a statistical theory of the sampling process. – Markov-Chains This is part of a new class of sampling methods termed adaptive/link-tracing designs (Thompson and Frank 2000) Anwendbar ist diese Methode allerdings nur. wenn überprüfbar ist. ob die potentiellen Interviewpartner tatsächlich der Untersuchungsgruppe angehören. also nicht nur. um eine Geldprämie zu erhalten. dies vorgeben. Institutional Sampling (Gewerkschaften. Outplacementberatung. ...) Time/ Space Sampling (Firma zum Zeitpunkt der Verlagerung) Targeted Sampling (Arbeitsämter. Transfergesellschaft) Chain-Referral Sampling (Convenience Sample – keine Rückschlüsse auf Grundgesamtheit)
  4. 06/07/11 RDS reduces the possibility of violating participant confidentiality during recruitment. Unlike the traditional chain-referral sampling . RDS does not ask participants to provide personal information about their peers; instead . it asks participants to recruit their peers. Importantly . it also helps reduce the “ masking effect . ” which means that respondents tend to protect their peers by not referring them in chain-referral sampling. Together with the financial reward for the referral effort . the rationing helps prevent the emergence of semi-professional recruiters and to reduce the effects of volunteerism . It is not necessary to have seeds randomly selected as long the seeds are members of the target population. No matter how the seeds are selected . the final RDS sample compositions will reach equilibrium (i.e. . stabilize) . independent of the characteristics of the seeds. Information about recruitment patterns in RDS can be used to estimate asymptotically unbiased population compositions using a mathematical model . called the reciprocity model . The estimated population compositions and sample compositions can be used create a sample weight . which is used to weight the sample. As such . unbiased sample statistics can be estimated from the weighted sample. This weight is defined as the inverse of the ratio of the group sample proportion to the estimate of corresponding population proportion. For example . suppose the gender compositions in an RDS sample were 60% male and 40% female; and the estimated corresponding gender compositions in the target population were 55% and 45% . respectively. This indicates that males were over-sampled and females were under-sampled. The weights used to account for male over-sampling and female under-sampling would be 1/(0.60/0.55) and 1/(0.40/0.45)=1.125 . respectively. The estimates of weights are available in the output of RDSAT. RDS sample analysis provides information for analysis of social structures. in which members of the target population are embedded.
  5. 06/07/11 The target population size needs to be large enough . theoretically . infinitely large . to apply RDS. That is . a reduction in population size due to earlier recruitment efforts will not influence the probabilities for recruiting the remaining population members. If the size of the target population is not large enough . RDS is not suitable. The reciprocity model used for population composition estimation in RDS sample analysis assumes reciprocal relationships between recruiters and recruitees. Heckathorn and colleagues (2002 . pp. 64) state that “In the absence of volunteerism and masking . recruits would be drawn randomly from respondents’ personal network.” As a matter of fact . social and geographic proximity among group members are likely to influence the way participants recruit from their personal networks. Information on personal network sizes . which is crucial for population composition estimation . is self-reported. To the best of our knowledge . reliability studies on the size and compositions of self-reported personal networks have never been conducted. Reliability studies on the size and compositions of self-reported personal networks have never been conducted. A good feature of RDS is that weights can be generated from the model-estimated unbiased population compositions together with sample compositions; unbiased statistics can . therefore . be estimated from the weighted sample. The RDS recruitment process is expected to grow geometrically (Heckathorn . 1997). Practically . however . this may not be the case.
  6. Survey was carried out from the end of December 2009 to March 2010 Three qualitative interviews . 96 former employees of that company could be identified. This number is similar to press releases about the relocation at that time . where 95 to 100 employees affected from the relocation were reported. 90 of those identified were eligible for the study purpose. From 37 former colleagues . none or wrong contact information could be found in the three preliminary interviews. This means only 53 questionnaires could be distributed among the former company members . which were filled out by 28 interviewees giving a response rate of 53 percent.
  7. Number of people identifiable through the network approach exceeds the number of eligible respondents identified through the three preliminary qualitative interviews Also show that there are bridges between people of various sub-groups in the network (i.e. between blue and white collar workers) Also shows is that there are still 15 isolated nodes in the network from which no path to another node in the component exists. While all 38 former white collar workers are part of the giant component (see also Table 1). 15 former blue collar workers are still missing. 14 of them are female with a high proportion of members from ethnic minorities (i.e.. 6 of them are Turkish. 4 are Yugoslav. 1 is Polish and only 3 are German).
  8. While the degree distribution in a random graph follows a bell curve. social networks often show a power law distribution. Most social networks are scale-free. Besides a high amount of people that have only one or two connections there are also a few nodes with up to 20 links to former colleagues. Those highly connected nodes are known as ‘hubs’ in network analysis. Hubs play a key role in keeping the network together (Ravasz & Barabási. 2003). which is especially important for the robustness of a shrinking network like those of former colleagues Additionally. hubs are also important for bridging different sub-populations in the network and will have important implications for the spread of information or rumors between those two sub-groups. Réka et al. (2000) demonstrated a high degree of robustness for scale-free networks. They show that even when as many as 5% of the nodes in a scale-free network are eliminated. “the communication between the remaining nodes in the network is unaffected” (Réka et al.. 2000. p. 380). High vulnerability to the removal of highly connected hubs in the network. for instance through death of a person with a high amount of links to other network members.
  9. z.B. würde die 73 die 38, 62 und 85 einladen
  10. Hinweis auf evtl. Couponverschwendung aber auch Peer-Pressure (Social Norm) z.B. Node 12 bekommt die Einladung sowohl von der 14 als auch von der 31 zu unterschiedlichen Zeitpunkten (Mehrfachteilnahme durch Speicherung der IP und diverser anderer Merkmale ausgeschlossen)