2017
DOI: 10.14301/llcs.v8i4.409
|View full text |Cite
|
Sign up to set email alerts
|

Comparing methods of classifying life courses: sequence analysis and latent class analysis

Abstract: We compare life course typology solutions generated by sequence analysis (SA)

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

1
32
0
3

Year Published

2018
2018
2021
2021

Publication Types

Select...
5
4

Relationship

0
9

Authors

Journals

citations
Cited by 36 publications
(36 citation statements)
references
References 49 publications
1
32
0
3
Order By: Relevance
“…It is opposed to conventional approaches in epidemiology that consider each state independently from one another. The comparison of SSA with other methods, which can be appropriate for the study of a care pathway, such as latent class analysis (LCA), revealed that both methodologies yield similar clustering results 41,49,50 . These studies have highlighted that SSA is easier to use and to compute than LCA and that, unlike LCA, it does not require specification of a model.…”
Section: Table 4 Here 4 Discussionmentioning
confidence: 99%
See 1 more Smart Citation
“…It is opposed to conventional approaches in epidemiology that consider each state independently from one another. The comparison of SSA with other methods, which can be appropriate for the study of a care pathway, such as latent class analysis (LCA), revealed that both methodologies yield similar clustering results 41,49,50 . These studies have highlighted that SSA is easier to use and to compute than LCA and that, unlike LCA, it does not require specification of a model.…”
Section: Table 4 Here 4 Discussionmentioning
confidence: 99%
“…We propose to use a combination of the two main criteria 38 : the weighted average silhouette width (ASWw) 38,39 and the Hubert's C index 40 allow for the optimal number of clusters to be set. The choice of these two parameters is in line with a recent social sciences paper, in which three criteria were selected for the partition 41 , namely the ASWw, Hubert's C index, and the point biserial correlation (PBC). We chose to forego the PBC as it usually led to the same choice of the number of clusters as the two other criteria.…”
Section: The 3 Rd Step Of Ssa: Clustering Sequencesmentioning
confidence: 99%
“…Next, we perform a hierarchical cluster analysis on this distance matrix using Ward's method, which implies that sequences with the smallest distance from each other are clustered. To determine the most appropriate number of clusters, we consider two cluster cutoff criteria-namely, the average silhouette widths (ASW) and point biserial correlation (PBC) (Studer 2013)-as well as the construct validity of the cluster solution, by comparing the fit of regressions of the different cluster-solutions on earnings-Akaike information criterion (AIC) and Bayesian information criterion (BIC) (Han et al 2017;Warren et al 2015). We use the TraMineR package in R to perform the sequence cluster analysis (Gabadinho et al 2011).…”
Section: Analytical Approachmentioning
confidence: 99%
“…Scientists can choose to prioritize indels to focus on duration in a state or substitution to focus on timing and order. Subsequence‐based metrics, as opposed to OM's edit‐based metrics, can also be used in sequence analyses, but OM has routinely outperformed when the life course sequences are related to family formation (Aisenbrey & Fasang, ; Han, Liefbroer, & Elzinga, ).…”
Section: Conceptualizing and Operationalizing Family Structure Historiesmentioning
confidence: 99%