Numeric Text Research Articles

Use of unstructured text in prognostic clinical prediction models: a systematic review.

ObjectiveThis systematic review aims to assess how information from unstructured text is used to develop and validate clinical prognostic prediction models. We summarize the prediction problems and methodological landscape and determine whether using text data in addition to more commonly used structured data improves the prediction performance.Materials and MethodsWe searched Embase, MEDLINE, Web of Science, and Google Scholar to identify studies that developed prognostic prediction models using information extracted from unstructured text in a data-driven manner, published in the period from January 2005 to March 2021. Data items were extracted, analyzed, and a meta-analysis of the model performance was carried out to assess the added value of text to structured-data models.ResultsWe identified 126 studies that described 145 clinical prediction problems. Combining text and structured data improved model performance, compared with using only text or only structured data. In these studies, a wide variety of dense and sparse numeric text representations were combined with both deep learning and more traditional machine learning methods. External validation, public availability, and attention for the explainability of the developed models were limited.ConclusionThe use of unstructured text in the development of prognostic prediction models has been found beneficial in addition to structured data in most studies. The text data are source of valuable information for prediction model development and should not be neglected. We suggest a future focus on explainability and external validation of the developed models, promoting robust and trustworthy prediction models in clinical practice.

AI in predicting COPD in the Canadian population

Chronic obstructive pulmonary disease (COPD) is a progressive lung disease that produces non-reversible airflow limitations. Approximately 10% of Canadians aged 35 years or older are living with COPD. Primary care is often the first contact an individual will have with the healthcare system providing acute care, chronic disease management, and services aimed at health maintenance. This study used Electronic Medical Record (EMR) data from primary care clinics in seven provinces across Canada to develop predictive models to identify COPD in the Canadian population. The comprehensive nature of this primary care EMR data containing structured numeric, categorical, hybrid, and unstructured text data, enables the predictive models to capture symptoms of COPD and discriminate it from diseases with similar symptoms. We applied two supervised machine learning models, a Multilayer Neural Networks (MLNN) model and an Extreme Gradient Boosting (XGB) to identify COPD patients. The XGB model achieved an accuracy of 86% in the test dataset compared to 83% achieved by the MLNN. Utilizing feature importance, we identified a set of key symptoms from the EMR for diagnosing COPD, which included medications, health conditions, risk factors, and patient age. Application of this XGB model to primary care structured EMR data can identify patients with COPD from others having similar chronic conditions for disease surveillance, and improve evidence-based care delivery.

Data Quality Considerations for Petrophysical Machine-Learning Models

Decades of subsurface exploration and characterization have led to the collation and storage of large volumes of well-related data. The amount of data gathered daily continues to grow rapidly as technology and recording methods improve. With the increasing adoption of machine-learning techniques in the subsurface domain, it is essential that the quality of the input data is carefully considered when working with these tools. If the input data are of poor quality, the impact on precision and accuracy of the prediction can be significant. Consequently, this can impact key decisions about the future of a well or a field. This study focuses on well-log data, which can be highly multidimensional, diverse, and stored in a variety of file formats. Well-log data exhibits key characteristics of big data: volume, variety, velocity, veracity, and value. Well data can include numeric values, text values, waveform data, image arrays, maps, and volumes. All of which can be indexed by time or depth in a regular or irregular way. A significant portion of time can be spent gathering data and quality checking it prior to carrying out petrophysical interpretations and applying machine-learning models. Well-log data can be affected by numerous issues causing a degradation in data quality. These include missing data ranging from single data points to entire curves, noisy data from tool-related issues, borehole washout, processing issues, incorrect environmental corrections, and mislabeled data. Having vast quantities of data does not mean it can all be passed into a machine-learning algorithm with the expectation that the resultant prediction is fit for purpose. It is essential that the most important and relevant data are passed into the model through appropriate feature selection techniques. Not only does this improve the quality of the prediction, but it also reduces computational time and can provide a better understanding of how the models reach their conclusion. This paper reviews data quality issues typically faced by petrophysicists when working with well-log data and deploying machine-learning models. This is achieved by first providing an overview of machine learning and big data within the petrophysical domain, followed by a review of the common well-log data issues, their impact on machine-learning algorithms, and methods for mitigating their influence.

Variable size sampling to support high uniformity confidence in sensor data streams

In order to rapidly process large amounts of sensor stream data, it is effective to extract and use samples that reflect the characteristics and patterns of the data stream well. In this article, we focus on improving the uniformity confidence of KSample, which has the characteristics of random sampling in the stream environment. For this, we first analyze the uniformity confidence of KSample and then derive two uniformity confidence degradation problems: (1) initial degradation, which rapidly decreases the uniformity confidence in the initial stage, and (2) continuous degradation, which gradually decreases the uniformity confidence in the later stages. We note that the initial degradation is caused by the sample range limitation and the past sample invariance, and the continuous degradation by the sampling range increase. For each problem, we present a corresponding solution, that is, we provide the sample range extension for sample range limitation, the past sample change for past sample invariance, and the use of UC-window for sampling range increase. By reflecting these solutions, we then propose a novel sampling method, named UC-KSample, which largely improves the uniformity confidence. Experimental results show that UC-KSample improves the uniformity confidence over KSample by 2.2 times on average, and it always keeps the uniformity confidence higher than the user-specified threshold. We also note that the sampling accuracy of UC-KSample is higher than that of KSample in both numeric sensor data and text data. The uniformity confidence is an important sampling metric in sensor data streams, and this is the first attempt to apply uniformity confidence to KSample. We believe that the proposed UC-KSample is an excellent approach that adopts an advantage of KSample, dynamic sampling over a fixed sampling ratio, while improving the uniformity confidence.

Integrating rich and heterogeneous information to design a ranking system for multiple products

The online review plays an important role as electronic word-of-mouth (eWOM) for potential consumers to make informed purchase decisions. However, the large number of reviews poses a considerable challenge because it is impossible for customers to read all of them for reference. Moreover, there are different types of online reviews with distinct features, such as numeric ratings, text descriptions, and comparative words, for example; such heterogeneous information leads to more complexity for customers. In this paper, we propose a method to integrate such rich and heterogeneous information. The integrated information can be classified into two categories: descriptive information and comparative information. The descriptive information consists of online opinions directly given by consumers using text sentiments and numeric ratings to describe one specific product. The comparative information comes from comparative sentences that are implicitly embedded in the reviews and online comparative votes that are explicitly provided by third-party websites to compare more than one product. Both descriptive information and comparative information are integrated into a digraph structure, from which an overall eWOM score for each product and a ranking of all products can be derived. We collect both descriptive and comparative information for three different categories of products (mobile phones, laptops, and digital cameras) during a period of 10days. The results demonstrate that our method can provide improved performance compared with those of existing product ranking methods. A ranking system based on our method is also provided that can help consumers to compare multiple products and make appropriate purchase decisions effortlessly.

Abstract P1-11-06: The experience of caregivers of women with metastatic breast cancer: Insights from the Make Your Dialogue Count survey

Abstract Introduction: Caregivers of patients with breast cancer have demonstrated persistent unmet needs, including reduced physical and psychosocial well-being. This may be particularly acute in caregivers of patients with metastatic breast cancer (MBC), whose ongoing treatments and increasing disability with disease progression offer particular challenges. While prevalence of MBC is currently unknown, caregivers of women with MBC represent a sizable group. Objective: To explore the emotional, psychological, and social impact on caregivers of patients with MBC and to identify gaps in communications between patients and their caregivers and oncologists. Methods: The "Make Your Dialogue Count" survey was conducted online, by paper, and by telephone (June-August 2014) among caregivers in the United States (age ≥18 y) who attended ≥50% of doctor visits of women with MBC (age ≥21 y). Survey responses were single- or multiple-response, numeric text, or rated on a 4-point Likert scale. Caregiver data were unweighted and representative only of those responding. Statistically significant differences between groups were determined by standard t-test of column proportions and means at the 95% confidence level. Results: 234 caregivers responded; 73% were men, 44% were the patient's spouse/partner, and 27% were nonwhite. Median age was 44 y. The caregivers' loved ones had a median time from initial breast cancer diagnosis of 44 months prior, and 51% had recurred after early breast cancer. Most caregivers (76%) considered themselves to be extremely or very involved in treatment decisions, but a sizable fraction of caregivers were unaware of the HR (20%) or HER2 (29%) status of the patient's cancer, indicating a lack of basic information needed for informed decision-making. While most (93%) said they felt comfortable speaking with the treatment team about MBC treatment, 41% reported communication barriers. Over half of caregivers (53%) felt that nobody understands what they're going through, and most (86%) reported that their lives had been negatively affected in some way (such as sleep habits, relationships and social life, hobbies and personal time, and financial stability and employment). Most caregivers considered caregiving to be an emotional (77%) and physical (56%) burden and 36% felt unappreciated. At the time of initial MBC diagnosis, 69% of caregivers felt it was important/very important for their loved one's doctor to refer them to support services, but only 25% of caregivers reported receiving such a referral. Subgroup analysis of caregiver's gender revealed some differences in responses relating to communications with oncologists, treatment experience, and emotional impact. Conclusions: While committed to their roles, these caregivers often found their role to be a physical and emotional burden, and many reported feeling isolated and unappreciated for their caregiving. Our findings indicate a strong need for support services specifically tailored to caregivers, including outreach to address emotional, financial, and practical needs stemming from caring for a loved one with MBC. They also indicate a need for improved disease and treatment information exchange between caregivers, patients, and healthcare providers. Citation Format: Mayer M, Sampayo I, Bell Dickson R, Citron ML, Brufsky AM. The experience of caregivers of women with metastatic breast cancer: Insights from the Make Your Dialogue Count survey. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-11-06.

An experiment testing six formats of 101-point rating scales

Although 101-point feeling thermometer questions are frequently used in population surveys and political polls, the measurement error associated with this question type has raised some concerns. One challenge of using this type of question arises from the tendency for respondents to round their answers. That is, rather than providing a precise answer, respondents provide an answer that is divisible by 5 which on at least some occasions is rounded up or down from a more precise (unrounded) value. Several national surveys currently use an open-ended numeric text input format for the feeling thermometer questions, which may exacerbate the response rounding. This study presents findings from two Web survey experiments that examine six formats of 101-point rating scales. The findings show that visual analog scales (VAS) result in fewer rounded answers than open-ended numeric text input, while item nonresponse rates are similar across these questions types. Also, responding to VAS tends to be easier than open-ended format. This study concludes by discussing the application of VAS for feeling thermometer questions and future research directions.

Understanding potential communication gaps between oncologists and patients with advanced breast cancer: A survey of oncologists.

124 Background: About 232,670 cases of breast cancer will be diagnosed in the US in 2014, of which 86,088 (37%) will be metastatic advanced breast cancer (ABC). The 2013 Global Count Us, Know Us, Join Us survey found that 53% of US women with ABC wanted more time to discuss their needs during clinic visits, and 60% believe their cancer treatment options are limited. To better understand unmet patient needs and potential communication gaps with physicians, surveys were developed for women with ABC, their caregivers, and oncologists. The oncologists’ survey results are presented. Methods: The Make Your Dialogue Count survey for oncologists has 32 items. Responses are single or multiple response numeric text or rated on a 4-point Likert scale. Harris Poll conducted this online survey from June through July 2014 among randomly selected, licensed US medical oncologists who treat ≥ 5 women with ABC per month. Responses for geographic region and number of years in practice by sex were weighted to match actual proportions in the population. Results: By July 15, 2014, 228 oncologists responded. At initial diagnosis, showing care and compassion was very important to 79% of oncologists; helping patients cope with diagnosis (62%) and manage cancer symptoms (56%) and discussing expected treatment response (57%) and side effects (51%) were also very important. Most oncologists (88%) revealed that treating women with ABC has a negative emotional toll. Although 81% of oncologists believe it to be unprofessional to let their emotion affect treatment recommendations, 26% reported that their emotions kept them from providing some information. Notably, 31% of oncologists reported that in certain situations, they have not discussed with patients that ABC is incurable. Conclusions: Caring for terminally ill patients has an emotional impact on oncologists. For some, their emotions have influenced what treatments they recommend and what information they impart. Our findings show that oncologists are compassionate and share disease- and treatment-related information and highlight the need to develop communication strategies that facilitate exchange between oncologists and their patients with ABC.

English

The aim of this study was to evaluate the clarity and consistency prevalent in dosing directions and measuring devices used for over the counter (OTC) liquid pediatric medications available in United Arab Emirates (UAE). 130 pediatric oral liquid OTC medications with dosing information for children younger than 12 years were studied. The study specifically focused on issues like the inclusion of a measuring device, child resistance packaging, within product inconsistency between dosing directions on the bottle’s label and dose markings on enclosed measuring device, across-product use of non standard units of measurements and abbreviations, use of numeric text according to the Food and Drug Administration recommendations (FDA) and the presence of  definitions for abbreviations used. Out of 130 preparations studied,  a measuring device was absent only in 16% (21 products, n=130). 31.5% (41 products, n=130) of the preparations studied did not have the child resistant cap.  Dosing directions on the label/leaflet and markings on the measuring device were the same in 53.6% (52 products, n=97) and different in 46% (45 products, n=97) of preparations. Superfluous markings (marking in the dosing device that is not referred to in the product labeled dosage instructions) were present in only 21.5% (23 products, n=107) of the cases. Inconsistent text for units (milliliters) was observed in 41.2% (40 products, n=97) products. Inconsistency in expressing teaspoon and tablespoon (inconsistency in expressing teaspoons as tsps or TSP and tablespoons as tbsp. or TBS) was 18.8% (13 products, n=69) and 28.5% (20 products, n=70), respectively. The use of non standard units, tablespoons and teaspoons needs to be quickly reviewed and prevented in the best interest of pediatric population.  Key words: Over the counter drugs, United Arab Emirates,  pharmaceutical preparations, oral drug administration, self administration.

A Review on Important Aspects of Information Retrieval

Information retrieval has become an important field of study and research under computer science due to explosive growth of information available in the form of full text, hypertext, administrative text, directory, numeric or bibliographic text. The research work is going on various aspects of information retrieval systems so as to improve its efficiency and reliability. This paper presents a comprehensive study, which discusses not only emergence and evolution of information retrieval but also includes different information retrieval models and some important aspects such as document representation, similarity measure and query expansion. Keywords—Information Retrieval, query expansion, similarity measure, query expansion, vector space model.

Using data sets in surgical pathology reporting of larynx carcinoma

It is imperative that anatomical pathologists report malignancy in a consistent, uniform method for many purposes including therapeutic management and database maintenance. Worldwide, surgical pathology reporting is shifting from the traditional text-based format, with specific information contained in the narrative, to synoptically structured communications. A synoptically structured report is a narrative report that is formally divided into explicit items covering specific observations on a specimen, and laid out in a predefined format, and may use Logical Observation Identifiers Names and Codes (LOINC). A synoptic report is one where the information has been fully separated into an explicit set of defined question and answer pairs of data items, and fully LOINC encoded for all questions, and for all answers that are not numeric values or free text entry. This standardisation of anatomical and microscopic information of malignancies facilitates the electronic transmission to cancer databases and registries for cancer surveillance. Both the North American Association of Central Cancer Registries (NAACCR) and the Royal College of Pathology (RCPath) have issued guidelines on synoptic reporting and an example for larynx carcinoma with brief reference to some of the important dataset criteria will be discussed.

Word-of-mouth and the forecasting of consumption enjoyment

The digital era has permitted rapid transfer of peer knowledge regarding products and services. In the present research, we explore the value of specific types of word-of-mouth information (numeric ratings and text commentary) for improving forecasts of consumption enjoyment. We present an anchoring-and-adjustment model in which the relative forecasting error associated with ratings and commentary depends on the extent to which consumer and reviewer have similar product-level preferences. To test our model, we present four experiments using a range of hedonic stimuli. Implications for the provision of consumer WOM are discussed.

OTC Liquid Medications Dosing and Devices: Confusion Reigns

Source: Yin H, Wolf M, Dreyer B, et al. Evaluation of consistency in dosing directions and measuring devices for pediatric nonprescription liquid medications. JAMA. 2010; 304(23): 2595– 2602; doi: 10.1001/jama.2010.1797Academicians from four United States medical schools studied dosing directions and measuring devices supplied with nonprescription liquid medications, determining conformity with Food and Drug Administration (FDA) guidelines published November 1, 2009.1 Guidelines were developed in response to reported unintentional drug overdoses ascribed to confusion generated by labels and measuring devices. Data reflect baseline adherence, against which the voluntary response by industry can subsequently be assessed.Utilizing an industry database (IMS Health Medicine Cabinet), medications for children younger than 12 years for the year ending on October 30, 2009 were determined. The 200 medications with the largest market share were chosen from five use categories: analgesic, cough/cold, combination analgesic and cough/cold, allergy, and gastrointestinal. Data abstracted included the presence and type of any enclosed measuring device, numeric dose amounts, exact text and abbreviations specifying units of dose measurement, format of decimals and fractions, and elements designed to guide consumers on appropriate use.Products were evaluated for differences between written dosing directions and the measuring device as well as for variability in units of measurement, abbreviations, format, and consumer guidance. Descriptive analyses were performed, including statistical association of findings with manufacturer type, medication category, and targeted age group. Products studied comprised 99% of the United States market share for nonprescription liquids. Most (74.0%) included a measuring device, usually a dosing cup (83.1%). A measuring device was supplied more frequently in products targeting infants and children exclusively (94.9% vs 60.7%) and with those supplied by large manufacturers (95.2% vs 46.3%). Provision varied by drug category: analgesics (95.5%); combination (92%); allergy (90.0%); cough/cold (71.1%); and gastrointestinal (50.0%). All associations are highly significant (P<.001). Measurement units differed between label and device in 89.0%. Devices had deficient (24.3%) or too many (81.1%) markings, and sometimes both. Some products (5.5%) used measures other than milliliters, teaspoons, or tablespoons. Many (65.0%) specified doses in multiple units. Nonstandard abbreviations were common. Dosing formats were potentially misleading and consumer guidance was rare.The authors conclude that at the time of the FDA’s new recommendations, the major pediatric OTC liquids had variable and inconsistent dosing instructions and measuring devices.Dr Ring has disclosed no financial relationship relevant to this commentary. This commentary does not contain a discussion of an unapproved/investigative use of a commercial product/device.Federal guidelines1 for dosing pediatric nonprescription liquid medications are evidenced-based, reasonable, and intuitive: provide a measuring device, with markings limited to those necessary for appropriate dosing; reconcile written dosing instructions with device markings; use only a few units (eg, milliliters, teaspoons, or tablespoons) on all packaging, labels, and devices; standardize and explain abbreviations; use only recommended formats for decimals and fractions; and provide consumers with key guidance (eg, use the device only for the product with which it is provided and clarify nonspecified doses with the prescriber). Using only metric measures2 might strengthen these recommendations. Additional factors3 need to be considered, including parental health literacy.In an accompanying editorial,2 a health communications expert considers this study in the context of the complete therapeutic encounter, emphasizing that safe and effective health care often depends on the patient (or caregiver) properly implementing treatment, especially in the outpatient setting. He opines that industry has greater responsibility for providing care related to nonprescription drugs and argues the necessity of increased research investment and more regulatory oversight.Yin and colleagues substantiate an important safety-related need4,5 for such guidelines and endorse the current federal effort. These baseline data clearly demonstrate that achieving conformity with the guidelines will require much work. Small manufacturers, cough/cold and allergy products, and products targeting broad age ranges may require particular scrutiny. Sourcing administration devices separately from the medications themselves may compound the problem. Continued focus on the drug categories and products with large market share is appropriate.The authors imply that stronger measures may be needed, including mandatory regulatory oversight in the absence of voluntary adherence by industry. Though providers are limited in how much they can influence the use of nonprescription drugs in the outpatient setting, they are nevertheless obligated to foster a culture of compliance based on clear, repetitive educational efforts. Patients and their caretakers must understand that accurate dosing and administration is important for prescription and nonprescription drugs alike. While industry and government can and should do more, limited FDA resources and increased emphasis on small government makes relying solely on industry and government insufficient.

Assessment and Improvement of Figures to Visually Convey Benefit and Risk of Stroke Thrombolysis

Deciding whether to use intravenous fibrinolytic therapy for acute cerebral ischemia within 3 hours of onset is challenging for patients, family members, and health care providers. Visual displays can permit individuals to rapidly understand response patterns to therapy. This study sought to evaluate, refine, and improve existing visual aids for stroke fibrinolytic decision-making. Existing visual aids were identified by Medline search and querying of national guideline organizations, pharmaceutical manufacturers, and stroke specialists, and were rated on a formal 8-point quality rating scale (0, lowest; 8, highest). Based on available instruments, new visual displays were developed to improve informed decision-making in routine practice. Two existing visual aids were identified, one from an emergency medicine society and one from a pharmaceutical company. Both were comparison visual displays of outcomes with and without treatment; no decision matrix visual aid was found. Both scored 4.0 on the quality scale, showing defects of effect size distortion, privileging less salient outcomes, dissimilar representation by treatment group, and limited stakeholder participation in generation. Revised versions of these graphics were developed with higher quality scores (6.75 and 7.75). In addition, a new decision matrix display with quality score 8.0 was developed that complements the numeric text of a national patient education tool developed jointly by US neurology, emergency medicine, and stroke patient organizations. Existing visual aids for stroke fibrinolysis decision-making have deficiencies. New visual displays are now available to convey the health benefits and risks of fibrinolytic stroke therapy efficiently and informatively to patients and family members.

Communicating side effect risks in a tamoxifen prophylaxis decision aid: The debiasing influence of pictographs

To experimentally test whether using pictographs (image matrices), incremental risk formats, and varied risk denominators would influence perceptions and comprehension of side effect risks in an online decision aid about prophylactic use of tamoxifen to prevent primary breast cancers. We recruited 631 women with elevated breast cancer risk from two healthcare organizations. Participants saw tailored estimates of the risks of 5 side effects: endometrial cancer, blood clotting, cataracts, hormonal symptoms, and sexual problems. Presentation format was randomly varied in a three factor design: (A) risk information was displayed either in pictographs or numeric text; (B) presentations either reported total risks with and without tamoxifen or highlighted the incremental risk most relevant for decision making; and (C) risk estimates used 100 or 1000 person denominators. Primary outcome measures included risk perceptions and gist knowledge. Incremental risk formats consistently lowered perceived risk of side effects but resulted in low knowledge when displayed by numeric text only. Adding pictographs, however, produced significantly higher comprehension levels. Pictographs make risk statistics easier to interpret, reducing biases associated with incremental risk presentations. Including graphs in risk communications is essential to support an informed treatment decision-making process.

Search across Different Media: Numeric Data Sets and Text Files

Digital technology encourages the hope of searching across and between different media forms (text, sound, image, numeric data). Topic searches are described in two different media: text files and socioeconomic numeric databases and also for transverse searching, whereby retrieved text is used to find topically related numeric data and vice versa. Direct transverse searching across different media is impossible. Descriptive metadata provide enabling infrastructure, but usually require mappings between different vocabularies and a search-term recommender system. Statistical association techniques and natural-language processing can help. Searches in socioeconomic numeric databases ordinarily require that place and time be specified.

An information extraction approach to reorganizing and summarizing specifications

Materials and Process Specifications are complex semi-structured documents containing numeric data, text, and images. This article describes a coarse-grain extraction technique to automatically reorganize and summarize spec content. Specifically, a strategy for semantic-markup, to capture content within a semantic ontology, relevant to semi-automatic extraction, has been developed and experimented with. The working prototypes were built in the context of Cohesia's existing software infrastructure, and use techniques from Information Extraction, XML technology, etc.

Presentation of real-time communication network measurement information

The presentation of packet-level communication network performance data to a user in an easy to assimilate, visual form by computer screen is considered. Standard presentation mechanisms such as bar graphs and numeric (text) displays require considerable concentration on behalf of the operator if more than a very small number of measurement values are presented. However, for many network operation tasks the initial requirement is for the relative value of the measurements and any trends in these measurements to be readily identifiable. A different approach to measurement presentation has therefore been considered whereby an object&apos;s shape is distorted in proportion to the values of a set of measurement data. A casual viewing of the object therefore immediately indicates the relative values of the measurements. This approach has been applied to the presentation of packet delay values for a commercially operational network, and its use in a performance measurement system for the commercial communication network is described.

Description and Prediction of Long-Term Learning of a Keyboarding Task

The goal of this study was to determine an equation (“learning function”) that describes long-term learning of a new keyboard. Five subjects learned 18 characters on a chord keyboard, then improved keying speed by inputting typical numeric keypad text for about 60 total hours. Their performance, in characters typed per minute, was recorded for every trial. Of the various functions that were considered to describe performance, the best fitting equation was a Log-Log relationship of the form CPMi = eb0Tib1, where CPMi is the performance in characters per minute on the i-th trial (Ti) and b0 and b1 are fitted coefficients. A second goal was to investigate how many trials of performance are needed before the entire learning function can be determined. The coefficients of the Log-Log function were determined using only the first 25, 50, 75, 100, 125, 150, 175, and 200 of the initial performance points (out of about 550 total actual data points). The mean squared error (MSE) was calculated for each of these fits and compared to the MSE of the fit using all points. From the results of MSE data, it appears that at least 50 performance data points are required to reduce the prediction error to an acceptable level.

The New Information Technologies and Psychiatry

The author reviews the history and technology of the microcomputer and discusses the various classes of software that are presently available. Three major categories of software are described: numeric data processing, text processing, and communications. The application of this software to psychiatric education and practice is briefly discussed. A short curriculum on computers for psychiatric residents is outlined, and a brief bibliography of the recent relevant literature on computer applications to medicine and psychiatry is presented. Predictions are made about the future direction of computer technology and its application to psychiatry.