On the Methods of Measuring Association Between Two Attributes

Yule, George

doi:10.2307/2340126

Cited by 629 publications

(288 citation statements)

References 4 publications

Supporting

Mentioning

271

Contrasting

Unclassified

Order By: Relevance

“…A convenient measure of association is Yule's (1912) coefficient of absolute association, V, which is closely related to x 2 in a 2 x 2 contingency table:…”

Section: T H E Incidence Of Non-viable# Organismsmentioning

confidence: 99%

An Outline of the Pattern of Bacterial Generation Times

Powell¹

1958

Journal of General Microbiology

145

View full text Add to dashboard Cite

SUMMARY: The generation times of four species of organisms have been measured, each under several sets of conditions : Aerobacter cloacae, Serratia marcescens, Streptococcus f aecalis and Pseudomonas aeruginosa. Minor variations in the experimental conditions appear to affect the mean generation time less in large samples than in small. This can be explained as a result of association between the generation times of closely related organisms. Positive correlation between the generation times of sisters, cousins and perhaps second cousins shows that the influence of an ancestor is felt through two or three generations. The observed correlation between mothers and daughters is usually small, probably because of bias due to the interval between fission of cytoplasm and fission of cell wall. The coefficient of variation of generation time is not a constant for the species but it is stable under given circumstances. It is possibly related systematically to the chemical complexity of the growth medium. In unhampered growth, less than 1% of the organisms produced are non-viable. There is positive association between the viabilities of sisters, and between the viability of an organism and the generation time of its mother. The distribution of generation times can be represented by a Pearson Type I11 or else a Pearson Type V distribution ; both are convenient in applications.The generation time of an individual is considered to be determined partly by molecular accidents, partly by heredity.In a previous paper on this subject (Powell, 1955), I gave an account of some preliminary work designed to test and compare the hypotheses of Rahn (1932) and Kendall (1948, 1952) about the causes of variability in the generation times ( 7 ) of individual bacteria. It turned out that a comparison of the observed distributions of 7 with those predicted by the hypotheses was not altogether satisfactory, and I suggested that a further test might be made by determining the effect of alterations in temperature and growth medium on the coefficient of variation of 7. At the same time I concluded that such experiments should await improvements in observational technique. There appeared to be a possibility that the relevance of the observations was altogether vitiated by ' delayed fission ', that is, by the lapse of an appreciable and variable interval between the termination of the processes considered by Rahn and Kendall, and the overt occurrence of fission.The results which I now present consist of generation time measurements on unicellular organisms of four species-Aerobacter cloacae, Serratia marcescens, Streptococcus faeculis and Pseudomonas aeruginosa-each under several different growth conditions. The undertaking was originally encouraged by the development of an improved culture chamber (Powell, 1956a), which would permit much greater accuracy in the determination of times of fission. At first, I hoped that the results would be of enhanced diacritical value, but it soon became clear that the generation times of bacteria do not depend ...

show abstract

“…A convenient measure of association is Yule's (1912) coefficient of absolute association, V, which is closely related to x 2 in a 2 x 2 contingency table:…”

Section: T H E Incidence Of Non-viable# Organismsmentioning

confidence: 99%

An Outline of the Pattern of Bacterial Generation Times

Powell¹

1958

Journal of General Microbiology

145

View full text Add to dashboard Cite

show abstract

“…As SVR does not provide the direction of the impact, this was estimated through the Phi coefficient ! test of association (Yule 1912). A positive Phi coefficient indicates a positive association between the independent variable and Reputation, while a negative Phi coefficient indicates a negative association between them.…”

mentioning

confidence: 99%

Who Am I? How Compelling Self-storytelling Builds Digital Personal Reputation

Pera

Viglia²,

Furlan

2016

Journal of Interactive Marketing

View full text Add to dashboard Cite

The work explores whether self-storytelling is a powerful predictor of personal reputation in a collaborative community of the sharing economy realm. By proposing that powerful self-storytelling allows an attractive positioning in respect to potential others, the paper extends the literature of brand storytelling and brand archetypes shifting the perspective to a personal level. This study adopts a qualitative-quantitative approach to investigate the meanings and stories contained in personal profile descriptions and their relation with reputation. Personal descriptions are interpreted as storytelling activities, labels/glosses that allow members to access the services of the community by facilitating personal reputation building. The findings show that powerful storytelling structures have defined phases and are crucial in reputation building when the story evolves in a metaphoric, symbolic lesson. The presence of archetypes, in particular the Sage and the Ruler, also confer reputational power to the stories. The results reveal opportunities for peer-to-peer communities, traditional companies, and social businesses. Marketers should design tools and platforms able to trigger consumers' desire to express their individuality through personal descriptions and suggest the drivers that affect reputation.Keywords: storytelling, personal reputation, archetypes, sharing economy 2 ! ! IntroductionA new, grassroots model of doing business is emerging, providing consumers with the power to get what they need and want where no transfer of ownership takes place. All over the world, people are renting rooms from strangers through Airbnb, outsourcing grocery trips to TaskRabbits, and traveling with ride-sharing service BlaBlaCar. These empowered individuals are participating in the sharing economy by beginning to function like hotels, taxis, farms, restaurants, manufacturers and other traditional business models. The sharing economy is becoming increasingly popular, and it is estimated, at the time we are writing, at more than US$15 billion (PwC 2014). The trend is expected to increase, as consumers and firms seek to maximize efficiency in volatile economic conditions (Lamberton and Rose 2012).The sharing economy is contingent upon one crucial factor: reputation, which is now considered a new currency for transactions in collaborative consumption platforms. Reputation is the enabling factor inherent within all sharing-sector activities. It helps to build trust in certain people and distrust in others. Because of its centrality to the success of the sharing economy, various thought leaders -entrepreneurs, social advocates, academics, investors, journalists -have opined as to how trust is established and maintained among strangers engaging in peer-to-peer transactions. Despite the topic has received attention, it still has not been theorized. This work explores personal reputation, understanding its antecedents. The main objective is to assess whether self-storytelling is a powerful predictor of personal reputation in o...

show abstract

“…Under statistical independence the value of the odds ratio is 1, but all other values of the odds ratio lie between 0 and ∞. Several authors have therefore proposed rescalings of the odds ratio that transform the measure to a correlationlike codomain (Yule, 1900(Yule, , 1912Digby, 1983). The association measures by Yule (1900Yule ( , 1912 and Digby (1983) have value 0 when two variables are statistically independent and maximum value 1 (perfect association).…”

Section: Rescaling the Odds Ratiomentioning

confidence: 99%

“…Several authors have therefore proposed rescalings of the odds ratio that transform the measure to a correlationlike codomain (Yule, 1900(Yule, , 1912Digby, 1983). The association measures by Yule (1900Yule ( , 1912 and Digby (1983) have value 0 when two variables are statistically independent and maximum value 1 (perfect association). Kraemer (1988) showed how 2 × 2 association measures like the sensitivity (a/p 1 ), specificity (d/q 1 ), predictive value of a positive Y (a/p 2 ), predictive value of a negative Y (d/q 2 ) and the four risk ratios can be transformed such that the new index has value 0 under statistical independence and a maximum value of 1.…”

Section: Rescaling the Odds Ratiomentioning

confidence: 99%

A Kraemer-type Rescaling that Transforms the Odds Ratio into the Weighted Kappa Coefficient

Warrens

2010

Psychometrika

View full text Add to dashboard Cite

This paper presents a simple rescaling of the odds ratio that transforms the association measure into the weighted kappa statistic for a 2 × 2 table.Key words: Cohen's kappa, 2 × 2 association measure. Measures of 2 × 2 AssociationIn a validity study a dichotomous variable Y is often compared to a 'gold standard' variable X. For example, in a medical test evaluation one has a 'gold standard' evaluation of the presence/absence or type of a disease against which a test is assessed. A 2 × 2 study can be summarized in a table like Table 1 (Warrens, 2008a(Warrens, , 2008b(Warrens, , 2009). In Table 1, the four proportions a, b, c, and d characterize the joint distribution of the variables X and Y . The row and column totals are the marginal distributions that result from summing the joint proportions. We denote these by p 1 and q 1 for variable X and by p 2 and q 2 for variable Y (Warrens, 2008c(Warrens, , 2008d.The odds ratio is a widely used measure of 2 × 2 association, and probably the most widely used measure in epidemiology (Edwards, 1963;Fleiss, 2003;Kraemer, 2004). The formula of the odds ratio in terms of proportions a, b, c, and d is OR = ad/bc. The odds ratio is the ratio of the odds of an event occurring in one group to the odds of it occurring in another group. These groups might be any other dichotomous classification. An odds ratio of 1 indicates that the condition or event under study is equally likely in both groups. An odds ratio greater than 1 indicates that the event is more likely in the first group.Another statistic of 2 × 2 association is the weighted kappa index (Spitzer, Cohen, Fleis, & Endicott, 1967;Vanbelle & Albert, 2009). It is the unique measure that is based on an acknowledgment that the clinical consequences of a false negative may be quite different from the clinical consequences of a false positive (Bloch & Kraemer, 1989;Kraemer, Periyakoil, & Noda, 2004). A real number r ∈ [0, 1] must be specified a priori indicating the relative importance of false negatives to false positives. The sample estimator of the weighted kappa (Bloch & Kraemer, 1989) isThe measure κ(1/2) is also known as Cohen's (1960) kappa (see also, Kraemer, 1979;Warrens, 2008e). Since the denominator of (1) can be written as Requests for reprints should be sent to Matthijs J. Warrens,

show abstract

On the Methods of Measuring Association Between Two Attributes

Cited by 629 publications

References 4 publications

An Outline of the Pattern of Bacterial Generation Times

An Outline of the Pattern of Bacterial Generation Times

Who Am I? How Compelling Self-storytelling Builds Digital Personal Reputation

A Kraemer-type Rescaling that Transforms the Odds Ratio into the Weighted Kappa Coefficient

Contact Info

Product

Resources

About