Isola Ajiferuke

Collaborative coefficient: A single measure of the degree of collaboration in research

It is shown that the mean number of authors per paper or the proportion of the multiple-authored papers is inadequate as a measure of the degree of collaboration in a discipline. A measure which combines some of the merits of both measures is suggested and derived. This measure, called the Collaborative Coefficient, is derived for four commonly used probability distributions.

Randomized controlled trials in pediatric complementary and alternative medicine: Where can they be found?

BackgroundThe safety and effectiveness of CAM interventions are of great relevance to pediatric health care providers. The objective of this study is to identify sources of reported randomized controlled trials (RCTs) in the field of pediatric complementary and alternative medicine (CAM).MethodsReports of RCTs were identified by searching Medline and 12 additional bibliographic databases and by reviewing the reference lists of previously identified pediatric CAM systematic reviews.ResultsWe identified 908 reports of RCTs that included children under 18 and investigated a CAM therapy. Since 1965, there has been a steady growth in the number of these trials that are being published. The four journals that published the most reported RCTs are The American Journal of Clinical Nutrition, Pediatrics, Journal of Pediatrics, and Lancet. Medline, CAB Health, and Embase were the best database sources for identifying these studies; they indexed 93.2%, 58.4% and 42.2 % respectively of the journals publishing reports of pediatric CAM RCTs.ConclusionsThose working or interested in the field of pediatric CAM should routinely search Medline, CAB Health and Embase for literature in the field. The four core journals identified above should be included in their collection.

Optimizing search strategies to identify randomized controlled trials in MEDLINE

BackgroundThe Cochrane Highly Sensitive Search Strategy (HSSS), which contains three phases, is widely used to identify Randomized Controlled Trials (RCTs) in MEDLINE. Lefebvre and Clarke suggest that reviewers might consider using four revisions of the HSSS. The objective of this study is to validate these four revisions: combining the free text terms volunteer, crossover, versus, and the Medical Subject Heading CROSS-OVER STUDIES with the top two phases of the HSSS, respectively.MethodsWe replicated the subject search for 61 Cochrane reviews. The included studies of each review that were indexed in MEDLINE were pooled together by review and then combined with the subject search and each of the four proposed search strategies, the top two phases of the HSSS, and all three phases of the HSSS. These retrievals were used to calculate the sensitivity and precision of each of the six search strategies for each review.ResultsAcross the 61 reviews, the search term versus combined with the top two phases of the HSSS was able to find 3 more included studies than the top two phases of the HSSS alone, or in combination with any of the other proposed search terms, but at the expense of missing 56 relevant articles that would be found if all three phases of the HSSS were used. The estimated time needed to finish a review is 1086 hours for all three phases of the HSSS, 823 hours for the strategy versus, 818 hours for the first two phases of the HSSS or any of the other three proposed strategies.ConclusionThis study shows that compared to the first two phases of the HSSS, adding the term versus to the top two phases of the HSSS balances the sensitivity and precision in the reviews studied here to some extent but the differences are very small. It is well known that missing relevant studies may result in bias in systematic reviews. Reviewers need to weigh the trade-offs when selecting the search strategies for identifying RCTs in MEDLINE.

A total relevance and document interaction effects model for the evaluation of information retrieval processes

Abstract The article presents a model based on the notion of the total relevance of a set of documents. The concept of a total relevance function is subsequently derived from the notion of cumulated relevance implied in the traditional summation of relevance ratings over the documents in a collection or in retrieved sets of documents. The model is intended to make explicit the perceptual underpinnings of relevance assessments while allowing for the consideration of interdocument dependencies as perceived by the user. Within this framework, it is proposed that an appropriate metric for gauging the performances of information retrieval systems is a measure of the (relative) total relevance that a user can obtain from a set of documents sequentially scanned and evaluated in an information retrieval environment. Some implications of the model are noted.

Citer analysis as a measure of research impact: library and information science as a case study

The investigators studied author research impact using the number of citers per publication an author’s research has been able to attract, as opposed to the more traditional measure of citations. A focus on citers provides a complementary measure of an author’s reach or influence in a field, whereas citations, although possibly numerous, may not reflect this reach, particularly if many citations are received from a small number of citers. In this exploratory study, Web of Science was used to tally citer and citation-based counts for 25 highly cited researchers in information studies in the United States and 26 highly cited researchers from the United Kingdom. Outcomes of the tallies based on several measures, including an introduced ch-index, were used to determine whether differences arise in author rankings when using citer-based versus citation-based counts. The findings indicate a strong correlation between some citation and citer-based measures, but not with others. The findings of the study have implications for the way authors’ research impact may be assessed.

Modelling count response variables in informetric studies: Comparison among count, linear, and lognormal regression models

The purpose of the study is to compare the performance of count regression models to those of linear and lognormal regression models in modelling count response variables in informetric studies. Identified count response variables in informetric studies include the number of authors, the number of references, the number of views, the number of downloads, and the number of citations received by an article. Also of a count nature are the number of links from and to a website. Data were collected from the United States Patent and Trademark Office (www.uspto.gov), an open access journal (www.informationr.net/ir/), Web of Science, and Macleans magazine. The datasets were then used to compare the performance of linear and lognormal regression models with those of Poisson, negative binomial, and generalized Poisson regression models. It was found that due to over-dispersion in most response variables, the negative binomial regression model often seems to be more appropriate for informetric datasets than the Poisson and generalized Poisson regression models. Also, the regression analyses showed that linear regression model predicted some negative values for five of the nine response variables modelled, and for all the response variables, it performed worse than both the negative binomial and lognormal regression models when either Akaikes Information Criterion (AIC) or Bayesian Information Criterion (BIC) was used as the measure of goodness of fit statistics. The negative binomial regression model performed significantly better than the lognormal regression model for four of the response variables while the lognormal regression model performed significantly better than the negative binomial regression model for two of the response variables but there was no significant difference in the performance of the two models for the remaining three response variables.

Searching for Controlled Trials of Complementary and Alternative Medicine: A Comparison of 15 Databases

This project aims to assess the utility of bibliographic databases beyond the three major ones (MEDLINE, EMBASE and Cochrane CENTRAL) for finding controlled trials of complementary and alternative medicine (CAM). Fifteen databases were searched to identify controlled clinical trials (CCTs) of CAM not also indexed in MEDLINE. Searches were conducted in May 2006 using the revised Cochrane highly sensitive search strategy (HSSS) and the PubMed CAM Subset. Yield of CAM trials per 100 records was determined, and databases were compared over a standardized period (2005). The Acudoc2 RCT, Acubriefs, Index to Chiropractic Literature (ICL) and Hom-Inform databases had the highest concentrations of non-MEDLINE records, with more than 100 non-MEDLINE records per 500. Other productive databases had ratios between 500 and 1500 records to 100 non-MEDLINE records—these were AMED, MANTIS, PsycINFO, CINAHL, Global Health and Alt HealthWatch. Five databases were found to be unproductive: AGRICOLA, CAIRSS, Datadiwan, Herb Research Foundation and IBIDS. Acudoc2 RCT yielded 100 CAM trials in the most recent 100 records screened. Acubriefs, AMED, Hom-Inform, MANTIS, PsycINFO and CINAHL had more than 25 CAM trials per 100 records screened. Global Health, ICL and Alt HealthWatch were below 25 in yield. There were 255 non-MEDLINE trials from eight databases in 2005, with only 10% indexed in more than one database. Yield varied greatly between databases; the most productive databases from both sampling methods were Acubriefs, Acudoc2 RCT, AMED and CINAHL. Low overlap between databases indicates comprehensive CAM literature searches will require multiple databases.

A probabilistic model for the distribution of authorships

A theoretical model for the distribution of authorships is developed. This model, the shifted Waring distribution, and 15 other discrete probability models are tested for goodness-of-fit against 94 data sets collected from six fields (engineering sciences, medical sciences, physical sciences, mathematical sciences, social sciences, and humanities). The shifted inverse Gaussian-Poisson is found to provide the best fitting.

Sample size and informetric model goodness-of-fit outcomes: a search engine log case study

The influence of sample size on informetric characteristics is examined to determine whether theoretical mathematical models can adequately fit large data sets. Two large data sets of queries submitted to the Excite search service were sampled for search characteristics (term frequencies, terms used per query, pages viewed per query, queries submitted per session) producing data sets of various sizes that were fitted to theoretical models to determine how the sample may influence a model’s goodness-of-fit. Although theoretical models could adequately fit smaller data sets of up to 5000 observations in some cases, larger data sets could not be satisfactorily fitted using several goodness-of-fit techniques. Investigators must take into account that sample size does influence goodness-of-fit outcomes. The nature of the data and not the limitations of given goodness-of-fit tests results in significant outcomes. Such goodness-of-fit tests should be used for comparative purposes, rather than significance testing.

Vocabulary size and its effect on topic representation

The impact of vocabulary reduction on topic modeling is explored for three data sets.Results are compared using four document and topic-centered measures.Removal of singly occurring terms has minimal influence on topics.Removal of frequently occurring terms greatly influences topic outcomes for three measures. This study investigates how computational overhead for topic model training may be reduced by selectively removing terms from the vocabulary of text corpora being modeled. We compare the impact of removing singly occurring terms, the top 0.5%, 1% and 5% most frequently occurring terms and both top 0.5% most frequent and singly occurring terms, along with changes in the number of topics modeled (10, 20, 30, 40, 50, 100) using three datasets. Four outcome measures are compared. The removal of singly occurring terms has little impact on outcomes for all of the measures tested. Document discriminative capacity, as measured by the document space density, is reduced by the removal of frequently occurring terms, but increases with higher numbers of topics. Vocabulary size does not greatly influence entropy, but entropy is affected by the number of topics. Finally, topic similarity, as measured by pairwise topic similarity and Jensen-Shannon divergence, decreases with the removal of frequent terms. The findings have implications for information science research in information retrieval and informetrics that makes use of topic modeling.