Latent structure facilitates development of research and theory

A latent variable is a variable that you can never observe directly. It can be inferred by statistical model by manifest (observable) variable. Being able to classify latent variable from observations, observations can be categorized in a new way which is rather unobservable.

For continuous manifest variable, the latent variable can be categorized (discrete groups) or continuous. Mixture model / Taxometric model can be used to find the categorized latent variable while factor analysis is used in general for continuous latent variable.

Implication

Latent structure facilitates development of research and theory. Distinct internet and Smartphone user groups (e.g. low, moderate and high addiction) were identified in each sex (Mok et al., 2014). The result showed that gender difference exists in internet and Smartphone usage and implied such addiction in male and female should be considered separately.

However, latent variable analysis should be used carefully. Taxonomy of life-course-persistent and adolescent-limited offending has been treated as categorized variable (Moffitt, 1993) which showed impact on research on delinquency and conduct disorder. It is recently found that the disorder may be continuous rather than distinct. Therefore, one should focus on investigating relative stability in delinquency dimensions than identifying distinct delinquency behavior (Piquero, 2008).

References

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: a developmental taxonomy. Psychological review, 100(4), 674.

Mok, J.-Y., Choi, S.-W., Kim, D.-J., Choi, J.-S., Lee, J., Ahn, H., . . . Song, W.-Y. (2014). Latent class analysis on internet and smartphone addiction in college students. Neuropsychiatric disease and treatment, 10, 817.

Piquero, A. R. (2008). Taking stock of developmental trajectories of criminal activity over the life course The long view of crime: A synthesis of longitudinal research (pp. 23-78): Springer.

Ross, C. A., Ellason, J. W., & Anderson, G. (1995). A factor analysis of the Dissociative Experiences Scale (DES) in dissociative identity disorder. Dissociation: Progress in the Dissociative Disorders.

Waller, N., Putnam, F. W., & Carlson, E. B. (1996). Types of dissociation and dissociative types: A taxometric analysis of dissociative experiences. Psychological Methods, 1(3), 300.

Latent Variable - Continuous or Categorized

A latent variable is a variable that you can never observe directly. It can be inferred by statistical model by manifest (observable) variable. Being able to classify latent variable from observations, observations can be categorized in a new way which is rather unobservable.

For continuous manifest variable, the latent variable can be categorized (discrete groups) or continuous. Mixture model / Taxometric model can be used to find the categorized latent variable while factor analysis is used in general for continuous latent variable.

Implication

Latent structure facilitates development of research and theory. Distinct internet and Smartphone user groups (e.g. low, moderate and high addiction) were identified in each sex (Mok et al., 2014). The result showed that gender difference exists in internet and Smartphone usage and implied such addiction in male and female should be considered separately.

However, latent variable analysis should be used carefully. Taxonomy of life-course-persistent and adolescent-limited offending has been treated as categorized variable (Moffitt, 1993) which showed impact on research on delinquency and conduct disorder. It is recently found that the disorder may be continuous rather than distinct. Therefore, one should focus on investigating relative stability in delinquency dimensions than identifying distinct delinquency behavior (Piquero, 2008).

References

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: a developmental taxonomy. Psychological review, 100(4), 674.

Mok, J.-Y., Choi, S.-W., Kim, D.-J., Choi, J.-S., Lee, J., Ahn, H., . . . Song, W.-Y. (2014). Latent class analysis on internet and smartphone addiction in college students. Neuropsychiatric disease and treatment, 10, 817.

Piquero, A. R. (2008). Taking stock of developmental trajectories of criminal activity over the life course The long view of crime: A synthesis of longitudinal research (pp. 23-78): Springer.

Ross, C. A., Ellason, J. W., & Anderson, G. (1995). A factor analysis of the Dissociative Experiences Scale (DES) in dissociative identity disorder. Dissociation: Progress in the Dissociative Disorders.

Waller, N., Putnam, F. W., & Carlson, E. B. (1996). Types of dissociation and dissociative types: A taxometric analysis of dissociative experiences. Psychological Methods, 1(3), 300.

種族歧視 - 關唔關我事?

最近有人指出係Google搜尋“three black teenagers” 同 “three white teenagers”，發現關於黑人的有很多嫌疑犯照片。相反，關於白人的則充滿歡樂。Google稱圖片的先後次序主要決定於在網上的出現次數。(http://atlantablackstar.com/2016/06/10/teen-googles-three-black-teenagers-and-three-white-teenagers-to-startling-results/)

可見，種族歧視十分普遍。但被問到會否歧視黑人時，我地通常都話唔會，而且真心覺得自己唔會有種族歧視的偏見。咁究竟點先知自己的取態呢?

內隱聯想測驗(Implicit Association Test，簡稱IAT)是一個於1998年由Greenwald Banaji同Nosek提出，用黎評估內在態度的測試。接受測試的人要跟據詞彙類別作出反應。例如詞彙同白人或正面的事有關，就按左鍵；如詞彙同黑人或負面的事有關，就按右鍵。此測驗原理為當兩個概念關係愈大(E.g. 白人、正面的事)，分辦該詞彙類別的反應就愈快。如果一個人對黑人的態度同對白人一樣，咁無論詞彙類別為 (白人 / 正面的事) VS (黑人 / 負面的事) 或(白人 / 負面的事) VS (黑人 / 正面的事)，反應速度都會一樣。

為左教育公眾IAT的概念，一班外國研究人員創辦了Project Implicit。呢個網站可以比大家測試自己對唔同事物的態度，如性別、性向、精神健康等等。有興趣了解自己隱藏的傾向可以去以下網站試試，結果會即時係完成測試後收到。

https://implicit.harvard.edu/implicit/

網上平台除左可以教育公眾，更可以同時收集世界各地的數據。此方法比發放問卷更容易，接受測試的人又可以即時知道結果，大大提升參與率。