A comparison of four computer-based telephone interviewing methods: Getting answers to sensitive questions

Interactive voice response (IVR) technology presents a new and promising approach by which to collect accurate data on sensitive topics by telephone interviews. In a national survey of 2, 880 households of alcohol and drug consumption, we compared computer-assisted telephone interviewing (CATI) and IVR with two hybrid methods that combine IVR with CATI. The principal hypothesis was that the self-report rates of sensitive behaviors would be higher for the hybrid and IVR methods owing to greater perceived confidentiality than with CATI. All the methods obtained similar sample demographic compositions. Response rates did not differ significantly between the CATI and the hybrid methods; however, the response rate with IVR was significantly lower. The hybrid and IVR methods obtained significantly higher self-report consumption rates for alcohol and marijuana and significantly higher hazardous drinking scores, as measured by the Alcohol Use Disorders Identification Test (AUDIT).     

A comparison of four methods of constructing factor scores

Four least-squares methods for constructing factor scores have been described in the literature. The formal properties of these scores are developed, and they are compared in terms of four generally desirable properties of constructed factor scores. In particular, it is shown that two of the methods yield scores that are conditionally unbiassed, and univocal in the sense of Guilford and Michael, though not orthogonal, while one of the other methods yields orthogonal scores.It is shown that constructed factor scores cannot be simultaneously univocal and orthogonal, unless we choose the special basis in factor space given by Canonical Factor Analysis.The general problem of choosing between the methods is discussed, on the basis of the theoretical relations obtained.EDITOR'S NOTE: The reader will quickly discover that this article develops several of the generalizations given in the second part of the preceding article, On Factors and Factor Scores. Independent development of the same generalizations is, of course, not a new phenomenon. Because the Presidential Address automatically is accepted for publication and given space in the December issue, it was decided that the only fair thing to do was to print this article in the same issue.     

A direct comparison of four methods for eliminating a response

Thirty-two rats pressed one lever (lever A) on a VI 30-sec schedule of food reinforcement and were then shifted to one of four procedures for eliminating the lever A response: extinction, differential reinforcement of other behavior, reinforcement of a different response (pole pushing), and reinforcement of a similar response (pressing lever B). Effectiveness of a response-elimination procedure was measured by (1) how quickly lever A response rate fell to a low level when the procedure was in effect, (2) how much lever A responding recovered when the procedure was discontinued, and (3) how resistant lever A responding was to reinstatement when the VI reinforcement schedule was reimposed. No one method was superior by all three measures. Extinction produced the most variable behavior, while differential reinforcement of other behavior produced the least. Reinforcing alternative behavior produced the greatest recovery in the original lever A response when the response-elimination procedure was discontinued.     

Response elimination: A comparison of four procedures

Four procedures for eliminating an operant response were compared within a multiple schedule. Response reduction was most rapid when reinforcement was provided for a specific alternative response. The decline in responding produced by extinction and differential reinforcement of other behavior was similar. Responding initially increased and then gradually decreased when reinforcers were delivered independently of responding at fixed times. Removal of reinforcement from the alternative-response and DRO procedures did not result in recovery of the target response. However, the shift from response-independent dilivery of reinforcers to extinction caused an initial increase in responding.     

How to handle missing data: A comparison of different approaches

Many researchers face the problem of missing data in longitudinal research. Especially, high risk samples are characterized by missing data which can complicate analyses and the interpretation of results. In the current study, our aim was to find the most optimal and best method to deal with the missing data in a specific study with many missing data on the outcome variable. Therefore, different techniques to handle missing data were evaluated, and a solution to efficiently handle substantial amounts of missing data was provided. A simulation study was conducted to determine the most optimal method to deal with the missing data. Results revealed that multiple imputation (MI) using predictive mean matching was the most optimal method with respect to lowest bias and the smallest confidence interval (CI) while maintaining power. Listwise deletion and last observation carried backward also scored acceptable with respect to bias; however, CIs were much larger and sample size almost halved using these methods. Longitudinal research in high risk samples could benefit from using MI in future research to handle missing data. The paper ends with a checklist for handling missing data.     

Detecting correlation changes in multivariate time series: A comparison of four non-parametric change point detection methods

Change point detection in multivariate time series is a complex task since next to the mean, the correlation structure of the monitored variables may also alter when change occurs. DeCon was recently developed to detect such changes in mean and\or correlation by combining a moving windows approach and robust PCA. However, in the literature, several other methods have been proposed that employ other non-parametric tools: E-divisive, Multirank, and KCP. Since these methods use different statistical approaches, two issues need to be tackled. First, applied researchers may find it hard to appraise the differences between the methods. Second, a direct comparison of the relative performance of all these methods for capturing change points signaling correlation changes is still lacking. Therefore, we present the basic principles behind DeCon, E-divisive, Multirank, and KCP and the corresponding algorithms, to make them more accessible to readers. We further compared their performance through extensive simulations using the settings of Bulteel et al. (Biological Psychology, 98 (1), 29-42, 2014) implying changes in mean and in correlation structure and those of Matteson and James (Journal of the American Statistical Association, 109 (505), 334-345, 2014) implying different numbers of (noise) variables. KCP emerged as the best method in almost all settings. However, in case of more than two noise variables, only DeCon performed adequately in detecting correlation changes.     

Conservatism as a predictor of responses to humour-I: A comparison of four scales

Four conservatism questionnaires (C-Scale; Wilson and Patterson, 1970; POI; Eysenck, 1976; MK; Cloetta, 1983; 16PF-Q1; Schneewind, Schröder and Cattell, 1983) were compared with regard to their ability to predict responses to humour based on the incongruity-resolution structure. We further investigated, whether a prediction of humour responses could be improved by assessing variables like toughmindedness, capitalism, rigidity and intolerance of ambiguity. The results showed that the conservatism scales overlap in their prediction of funniness of incongruity-resolution jokes, sex jokes and rejection of nonsense jokes. Correlations with the other humour scales were specific for certain conservatism questionnaires. A canonical correlation analysis relating the humour scores and the conservatism scales yielded three significant correlations (r_c1 = 0.65, r_c2 = 0.51 and r_c3 = 0.44). Capitalism was also related to appreciation of incongruity-resolution jokes and of sex jokes. The previously found effects of intolerance of ambiguity could be replicated.     

Comment on “four methods of locating the egocenter: A comparison of their predictive validities and reliabilities”

Barbeito and Ono (1979) recently reported a psychometric evaluation of four methods of measuring the egocenter. Evaluations of this kind can be extended, with a view to obtaining independent estimates of reliability and validity coefficients. In this way, the advantage(s) of one type of measure over another can be more clearly delineated.     

A comparison of analysis methods to estimate contingency strength

To date, several data analysis methods have been used to estimate contingency strength, yet few studies have compared these methods directly. To compare the relative precision and sensitivity of four analysis methods (i.e., exhaustive event‐based, nonexhaustive event‐based, concurrent interval, concurrent+lag interval), we applied all methods to a simulated data set in which several response‐dependent and response‐independent schedules of reinforcement were programmed. We evaluated the degree to which contingency strength estimates produced from each method (a) corresponded with expected values for response‐dependent schedules and (b) showed sensitivity to parametric manipulations of response‐independent reinforcement. Results indicated both event‐based methods produced contingency strength estimates that aligned with expected values for response‐dependent schedules, but differed in sensitivity to response‐independent reinforcement. The precision of interval‐based methods varied by analysis method (concurrent vs. concurrent+lag) and schedule type (continuous vs. partial), and showed similar sensitivities to response‐independent reinforcement. Recommendations and considerations for measuring contingencies are identified.     

A comparison of methods to teach foreign-language targets to young children

Using instructional strategies based on derived relational responding (DRR) to teach foreign-language targets may result in emergent, untrained foreign-language relations. One benefit of using DRR instructional strategies is the efficiency with which an individual acquires additional stimulus relations as a result of emergent responding following acquisition of one or a small number of relations. In the current study, we compared the efficiency of tact training alone to a traditional foreign-language teaching strategy (i.e., teaching all relations concurrently—mixed training) with four, 4-year-old children. The results demonstrated that tact training was more efficient than mixed training for 5 of 7 stimulus sets. The findings add to the research demonstrating that DRR instructional strategies, specifically tact training, may be more efficient than concurrently teaching all targeted relations.     

A comparison of four measures of self-control skills

This study compares the psychometric characteristics of four questionnaires designed to assess self-control skills: the Self-Control Questionnaire, the Frequency of Self-Reinforcement Questionnaire, the Cognitive Self-Management Test, and the Lifestyle Approaches Inventory. Content validity was judged to be fairly comparable by three raters in Study 1. In Study 2, convergent and divergent validity support was obtained for all four questionnaires when administered to 369 multiethnic college students, but the relative degree of support varied across constructs. Hence, selection of a self-control instrument may be guided by the target behavior of interest.     

Visual regulation of manual aiming: A comparison of methods

Visual regulation of upper limb movements occurs throughout the trajectory and is not confined to discrete control in the target area. Early control is based on the dynamic relationship between the limb, the target, and the environment. Despite robust outcome differences between protocols involving visual manipulations, it remains difficult to identify the kinematic events that characterize these differences. In this study, participants performed manual aiming movements with and without vision. We compared several traditional approaches to movement analysis with two new methods of quantifying online limb regulation. As expected, participants undershot the target and their movement endpoints were more variable when vision was not available. Although traditional measures such as reaction time, time after peak velocity, and the presence of discontinuities in acceleration were sensitive to the visual manipulation, measures quantifying the trial-to-trial spatial variability throughout the trajectory were the most effective in isolating the time course of online regulation.     

A comparison of recognition memory to numerical decision: How prior probabilities affect cutoff location

In two experiments, a recognition memory task, in which subjects judged whether a given five-digit number had been shown previously, was compared to a numerical decision task, in which subjects judged whether a given five-digit number represented the height of a man or a woman. Subjects were found to shift β (the cutoff point along the decision axis) with changes in prior probabilities in the direction specified by signal detection theory in the numerical decision task but not in the recognition memory task. An explanation of the results, in terms of a difference in ? (discriminability) between the two tasks, was ruled out. In contrast, explanations which cannot be ruled out involve differences between the tasks in the amount of practice required for the subject to learn the manner in which the stimuli are distributed along the decision continuum, and differences in the availability of an alternative basis for response other than the decision continuum.     

A comparison of six methods to estimate thresholds from psychometric functions

There are many ways in which to estimate thresholds from psychometric functions. However, almost nothing is known about the relationships between these estimates. In the present experiment, Monte Carlo techniques were used to compare psychometric thresholds obtained using six methods. Three psychometric functions were simulated using Naka-Rushton and Weibull functions and a probit/logit function combination. Thresholds were estimated using probit, logit, and normit analyses and least-squares regressions of untransformed orz-score and logit-transformed probabilities versus stimulus strength. Histograms were derived from 100 thresholds using each of the six methods for various sampling strategies of each psychometric function. Thresholds from probit, logit, and normit analyses were remarkably similar. Thresholds fromz-score- and logit-transformed regressions were more variable, and linear regression produced biased threshold estimates under some circumstances. Considering the similarity of thresholds, the speed of computation, and the ease of implementation, logit and normit analyses provide effective alternatives to the current “gold standard”—probit analysis—for the estimation of psychometric thresholds.     

A review and empirical comparison of motivation scoring methods: An application to self-determination theory

Motivation and Emotion - Self-determination Theory differentiates various types of motivation, each of which have different consequences for well-being and behavior. Despite broad agreement...