Association for Behavior Analysis International

The Association for Behavior Analysis International® (ABAI) is a nonprofit membership organization with the mission to contribute to the well-being of society by developing, enhancing, and supporting the growth and vitality of the science of behavior analysis through research, education, and practice.


40th Annual Convention; Chicago, IL; 2014

Event Details

Previous Page


Poster Session #94
Saturday, May 24, 2014
5:00 PM–7:00 PM
W375a-d (McCormick Place Convention Center)
23. Cross-Cultural Challenges to University Collaboration: Creating Sustainable Training in ABA
Area: EDC; Domain: Service Delivery
MARY E. BRADY (University of Massachusetts Boston), Pooja Panesar (Kaizora Consultants), Molly Ola Pinney (Global Autism Project)
Abstract: Interest in global outreach from countries with well-establish Applied Behavior Analysis services and training programs is evident. Check university websites promoting international service learning opportunities as a component of their ABA degree programs, and direct service agencies posting photos of good deeds done with underserved children in other countries. Our ABA community has enough introductory global outreach experience to reflect upon cross-cultural challenges and social/ethical concerns arising when we do “good work” from a power position based upon knowledge and resources. Specifically, we’ll focus upon cross cultural challenges universities experience when providing ABA training internationally, including: Relationships with partner universities with different protocol and pace, and frequent changes in faculty and administrators that fracture progress, Financial differences in tuition costs and income differentiation, raising ethical and moral concerns that we are educating only the elite, Inconsistent access to more affordable social media and technology tools that provide and maintain supervision and knowledge sharing, Training sites for supervised experience, Social environments (violence, strikes, political turmoil especially surrounding elections, Finally, we will recommend a model to establish global university training, one component of a sustainable service delivery system for children with autism in countries with an emerging awareness of their needs.
24. Reducing College Students’ Procrastination: A Review and Future Research Suggestions
Area: EDC; Domain: Applied Research
ZIWEI XU (The Ohio State University), Marnie Nicole Shapiro (The Ohio State University), Nancy A. Neef (The Ohio State University)
Abstract: Procrastination is a maladaptive study pattern frequently observed among college students which may result in physical, emotional, and social problems. In behavior analytic research, procrastination is often a result of an ineffective natural contingency. That is, individuals are more likely to engage in avoidance behavior when a deadline is distant and the consequences of their responses are too minimal to reinforce study behaviors. Since the early 1970s, researchers have focused on the development of competing contingencies to reduce college students’ procrastination that commonly include self-management techniques and instructor-administered contingencies. Although the procedures have been effective in reducing procrastination, our review of the literature also indicated that the resultant impact on students’ academic performance is unclear. In this poster, we review and discuss evidence-based strategies to reduce procrastination, as well as the collateral effects on students’ academic performance. In addition, we will discuss 1) implications of our findings as they relate to effective college teaching, 2) suggestions for future research, and 3) challenges for both practice and research in the area of reducing student procrastination.
25. The Prevalence of the Overjustification Effect in Undergraduate Students when Rewards are Expected vs. Unexpected
Area: EDC; Domain: Applied Research
ZACH TROUTMAN (Penn State University), Jeremy DaShiell (Pennsylvania State University - Harrisburg), Julie A. Ackerlund Brandt (Penn State Harrisburg)

Despite the efficacy of reinforcement-based procedures for increasing appropriate behaviors, there is still reluctance to adopt them, especially within educational settings. This reluctance is due to a popular Social Theory concept called the Overjustification Effect. Researchers have suggested that reinforcement procedures reduce intrinsic motivation to perform a behavior after reinforcement has been terminated. The purpose of the present experiment was to determine the prevalence and magnitude of the overjustification effect when token reinforcement is withheld following a period of reinforcement. There have been additional suggestions that the expectancy of rewards increases the likelihood of the overjustification effect; therefore, this study also compared expected and unexpected reinforcement conditions. A reversal design was used to analyze the experimental conditions in which undergraduate students were presented with basic math problems under different reinforcement contingencies. The overjustification effect was observed in 2 of the 12 participants (once in the unexpected condition; once in the expected condition). Overall, the results of the study did not show a significant prevalence of the overjustification effect, and there does not appear to be a difference as to the expectancy of rewards.

26. The Relationship Between Completion of Quizzes, Impulsivity Scores, and Estimated Study Time
Area: EDC; Domain: Basic Research
DEBRA J. SPEAR (South Dakota State University)
Abstract: As part of the course requirement, Honors students in General Psychology could complete 10 graded chapter quizzes, available online throughout the semester. The grade and the time the quiz was completed were recorded. For their research requirement, students completed two survey measures of impulsive behavior (BIS, AISS), basic demographic questions, a brief hypothetical gambling task, and estimated study times in six different scenarios. In the six hypothetical situations described, three factors varied: at the beginning/middle of the semester, that was in/not in their major, either a lot of work or little work was required. Students estimated how much time they would put into the course if they knew that students in the past put in X hrs. and received an A (25, 50, 75, 100 or 125 hrs.) or Y hrs. and received a C (5, 15, 25., 50, 75, or 100 hrs). There was a weak correlation between BIS score and the average amount of time a student waited to complete an online quiz. The same was true for the correlation between quiz wait time and AISS scores (Overall, Impulsivity Subscale, and Strength Subscale). As the number of hours required to obtain an A increased, all students increased the estimated amount of time they would devote to the hypothetical class. However, as the A required increasingly more time, students were slightly less likely to devote the maximum hours to the course the later it was in the semester and if it was a non-major class.
27. The Effects of Lighting on Student Behavior - an Experimental Analysis.
Area: EDC; Domain: Applied Research
RYAN SAIN (Eastern Washington University), Benjamin Kowal (University of Arkansas at Little Rock), Mark Lee Olsen (Eastern Washington University)
Abstract: In Hawthorne's classic study, he simply manipulated lighting conditions. This study sought to replicate his findings with a more thorough investigation of the effects of lighting on the college classroom. Several behaviors of several individuals were observed using multiple observers. Specifically, attentive vs. inattentive behavior and three ratings of the participants' posture were recorded using a two minute momentary time sampling technique for a duration of 50 minutes, four times a week for a total of 10 weeks. A full reversal design was employed. Additionally, the study was also conducted as a multiple baseline across different university classes (and individuals). In the baseline conditions lights were on and in the intervention conditions lights were off. It was not possible to control for natural lighting effects, however both classrooms did have some natural light. Results indicate that there was a mild effect of lighting on attentiveness and posture. Implications of the findings are discussed in the context of effective teaching practices in the college classroom.
28. Cumulative Record of Student Exam Attempts in a Mastery Based University Course
Area: EDC; Domain: Applied Research
RYAN SAIN (Eastern Washington University), Benjamin Kowal (University of Arkansas at Little Rock), Lan Tran (Eastern Washington University)
Abstract: Historically, fixed interval schedules of reinforcement have been associated with a post-reinforcer pause. However, some have observed that this pause is actually before the behavioral run and have thus identified it as a procrastination pause. The current study uses a single subject approach to evaluate how students responded to changes in a limited hold attached to the fixed interval schedule of completing online exams in a university course. The cumulative record of student attempts at the exams is presented. It is clear that few students are behaving in a traditional, post-reinforcer pause fashion and most are behaving in a procrastination pause form. Students did, however, respond clearly to the changes in the limited hold that was systematically varied across the academic term. It also seems that the number of test attempts is not related to the likelihood of the procrastination or post-reinforcer pause. Further, when exams were opened concurrently there seemed to be a clear pause from the completion time of one to the beginning of another (evidence for post-reinforcer pauses). It is still not clear then what variables control the behavior of procrastination vs. post-reinforcer pauses. Implications for this finding are discussed in detail.
29. Within-Session Changes in Attentive Behavior Among College Students During Class
Area: EDC; Domain: Applied Research
BENJAMIN KOWAL (University of Arkansas at Little Rock), Ryan Sain (Eastern Washington University), Mark Lee Olsen (Eastern Washington University), Justine Recor (Eastern Washington University)
Abstract: When responding is observed within sessions of operant reinforcement, rates of responding typically increase, decrease, or increase and then decrease. Within-session patterns suggest caution in comparing average responses rates across conditions and under certain conditions may help explain preferences. In the current investigation, attentive and inattentive behaviors of college students were observed in a classroom setting to determine if within-session patterns were present for each type of behavior and whether within-session patterns might explain preferences for attentive behavior. Student behavior was monitored and coded using a momentary time sampling procedure every two minutes by an observer, during daily 50 min lectures, over the course of an entire quarter. Attentive behavior as well as the proportion of attentive relative to inattentive behavior (i.e., the operational definition of preference used in this study) in each interval tended to increase within classroom sessions. Inattentive behavior tended to increase at the beginning of the session and subsequently decrease, particularly towards the second half of the class. Within-session changes in attentive and inattentive classroom behavior observed in the current study fit in well with the previous literature and may, at least in part, explain when students prefer to engage in inattentive or attentive behaviors.
30. Enhancing Traditional Medical Education (and the Resulting Clinical Care) using Behaviorally Based Principles
Area: EDC; Domain: Applied Research
Abstract: The time-honored mantra of training medical students and residents, "See one! Do one! Teach one!," has a catchy sound, but often isn't adequate for the complex and even not so complex procedures and practices of clinical medicine. It is especially inadequate for teaching generalization to the myriad circumstances, and human tragedies, of day to day clinical medicine. It is time honored, but more and more readily seen as inefficient and unsafe. Integrating basic principles of behaviorally based education into the traditional practices of medical education can be challenging in that it must overcome several hundred years of learning history and traditional practices of those teaching. Examples of behaviorally based education include direct supervision in real time of procedures by students and junior residents, greater emphasis of teaching basic skills in preclinical years before they are expected to be performed in clinical clerkships on "real patients," the widespread use of technologically sophisticated simulation labs, and the particularly behavioral yet often not done practice of actually deconstructing a complex procedural skill into component skills and behavioral objectives, and then practicing these component behavioral chains until the basic pattern is learned well, and then generalized by gradually altering the practice circumstances including increasingly complex variables and decision trees. Facilitating attempts to implement more behaviorally based teaching strategies for healthcare providers creates a culture of higher expectations. This talk reviews examples of behaviorally sound teaching in the classroom, clinic, and bedside, as well approaches to changing habits of medical educators.
31. Examining the Components of Online Interteaching
Area: EDC; Domain: Applied Research
ANITA LI (Florida Institute of Technology), Joshua K. Pritchard (Florida Institute of Technology)
Abstract: Interteaching is a method of instruction that has been demonstrated as a method to enhance college instruction beyond the traditional lecture format. While interteaching has been demonstrated effective in live face-to-face classrooms, it is time to move it online! We will discuss the process and impact when we move interteaching online specifically in relation to its effect in online behavior analytic coursework and BACB exam preparation. A preliminary experiment was conducted by the authors to determine the effectiveness of interteaching online which had promising results. We will describe and discus the ideal conditions for the next generation in interteaching -- online!
32. An Evaluation of Computer-Based Instruction on the Visual Analysis of Single-Subject Data
Area: EDC; Domain: Applied Research
KATIE WOLFE (University of South Carolina), Timothy A. Slocum (Utah State University)
Abstract: Visual analysis is the primary method of analyzing data in single-subject methodology. Previous research on the reliability of visual analysis suggests that judges often disagree about what constitutes an intervention effect. Given that visual analysis involves complex discriminations and sometimes produces disagreement among experts, it is important to examine methods of training individuals to visually analyze data. The purpose of this study was to evaluate two training methods using graphs with various combinations of slope and level changes. The computer-based training, which includes a very high number of practice opportunities with feedback, was compared to a lecture condition and a control condition. Results indicate that both training methods were more effective than a control condition, but were not substantially different from one another. We discuss the implications of these results for training individuals in visual analysis as well as directions for future research.
33. Stimulus Equivalence and the Emergence of Topography-Based Driving Behavior
Area: EDC; Domain: Applied Research
ANDREW BLOWERS (Southern Illinois University Rehabilitation Institute), Ruth Anne Rehfeldt (Southern Illinois University)
Abstract: The acquisition of behaviors required to operate an automobile can be viewed as a behavioral cusp since such behavioral change would facilitate an environment in which an individuals repertoire would contact new contingencies resulting in pervasive consequential events. Three individuals with learning disabilities and no state drivers license were exposed to a selection-based stimulus equivalence protocol via a computer program. The selection-based instructional protocol was intended to teach participants to identify correct relations between stimuli associated with automobile operation. Following pretest, participants were directly taught relations between video models of appropriate in-car behaviors, outlines of road signs, and textual stimuli contained within a given road sign. The emergence of untaught relational responding was evaluated at posttest once participants met mastery for directly taught relations. Generalization probes evaluated the emergence of topography-based responding on a driving simulator task by presenting participants with point-of-view video clips and asking them to respond as they would if operating a vehicle in the natural environment. A selection-based instructional protocol that results in the emergence of complex topography-based responding, such as behaviors required to operate a vehicle, has widespread implications. Suggestions for future research are provided.
34. The Impact of Applied Behavior Analysis on Law Enforcement Video Simulation Training
Area: EDC; Domain: Applied Research
JAMES MEADOR (Graduate student), Kent Corso (NCR Behavioral Health, LLC)
Abstract: Law enforcement officers (LEOs) routinely face lethal situations in the field which require near instantaneous responses. To minimize the risk of death to LEOs or civilians, LEO field training must be effective. The law enforcement (LE) field generally deems video simulation training (VST) as a gold standard for teaching officers to handle lethal situations, including officers appropriate use of force. Yet, no known behavior analytic data have been published on the effectiveness of these VST modules. The training relies on video contingencies which LEOs interact with via role play, along with use of rule-governed contingencies delivered in didactic format. This is the first known exploratory program outcome evaluation of LE VST using applied behavior analysis. The authors examined differences in officers demonstration of skills and latency of engaging with the simulated contingencies by randomly assigning each officer to one of two conditions: modelling (N=39) and training as usual (TAU; N=33). Results were analyzed via scatterplot. The level and variability of responses were higher in the TAU condition, indicating delayed engagement with the VST modules. Similarly, rates of correctly performing several skills were higher in the modeling condition. Implications and future directions for military and LE training are discussed.
35. The Effects Of Single Versus Multiple Exemplar Training On Vocal Identification Of Artists' Styles
Area: EDC; Domain: Applied Research
MARCHELE TUCKER (The Chicago School of Professional Psychology), Jessica Gamba (The Chicago School of Professional Psychology), Diana J. Walker (The Chicago School of Professional Psychology), Fawna Stockwell (The Chicago School of Professional Psychology)
Abstract: The purpose of this study was to assess the effects of single versus multiple exemplar training on graduate students’ vocal identification of artists’ styles and stimulus generalization to novel paintings by the same artists. Six graduate students over the age of 18 participated in this study. The study utilized decks of cards that depicted images of paintings by six different artists. Each condition was implemented with one of two sets of artists. Multiple exemplar decks had a total of nine cards, with three exemplars of each artist’s paintings. The single exemplar decks had three cards, one exemplar of each artist's paintings. Overall, five of six participants performed better on the generalization test after multiple exemplar training compared to single exemplar training. The mean percentage of correct responses on the generalization test after multiple exemplar training was 86.7% (range, 80% to 100%). After single exemplar training the mean percentage correct was 76.7% (range, 46.7% to 100%). These results suggest that multiple exemplars of each artist’s style improved generalization to the novel paintings from each artist compared to training with a single exemplar from each artist.
36. The Effects of Voluntary versus Cold-calling Participation on Class Discussion and Performance Measures in an Undergraduate Course
Area: EDC; Domain: Basic Research
BRITTANY A. CARSTENS (University of Tennessee), Tiffany Best (University of Tennessee), Eleanore Claire Trant (The University of Tennessee), Katie Crabtree (University of Tennessee), Leslie Hart (University of Tennessee), Samantha Adair (University of Tennessee), Carrie Jaquett (University of Tennessee), Robert Lee Williams (The University of Tennessee)
Abstract: Although class participation has been linked to improved student performance, little research has evaluated the effects of cold-calling versus voluntary participation. This study aims to determine the differential effects of voluntary and cold-calling participation practices on a variety of participation and performance variables in a college setting. Approximately 200 students from eight sections of an undergraduate course involving extensive instructor-led discussion will serve as participants. Instructors of each section will alternate the use of voluntary and cold-calling participation across units in the semester-long course. The effects of voluntary and cold-calling conditions on attendance, individual participation, and exam performance across units will be examined within units across class sections. Results will be evaluated by comparing the voluntary and cold-calling conditions for combined and individual sections, with baseline values in each section used as the covariate. Preliminary results indicate students introduced to the cold-calling condition early during the course have significantly lower voluntary participation levels than students introduced to the cold-calling condition. Further findings will be presented.
37. The Effects of Unstructured Laptop Use Within a Highly Structured Undergraduate Class
Area: EDC; Domain: Applied Research
BRITTANY A. CARSTENS (University of Tennessee), Tiffany Best (University of Tennessee), Eleanore Claire Trant (The University of Tennessee), Jennifer Wright (University of Tennessee), Jeremy Coles (University of Tennessee), Robert Lee Williams (The University of Tennessee)
Abstract: The current study examined the effects of unstructured laptop use on levels of class participation and exam performance in multiple sections of a highly structured undergraduate course. Students were permitted to use their laptops during specified course units. Survey data were collected at the conclusion of the laptop units to determine students’ self-reported frequency of laptop use for a variety of purposes. We used a three-way mixed design in determining main and interaction effects between laptop available vs. non-available course units, later vs. early laptop-available units in the course, and high, medium, or low laptop use in laptop-available units. Results showed that in a highly structured class, students participated more in class discussion when allowed to use their laptops. In contrast to its effect on class participation, laptop availability did not significantly affect exam performance. Survey results indicated a majority of students preferred not to use their laptops during class, and those who preferred laptop use primarily used their laptop for taking notes and examining course content.
38. Behavior Oriented Analysis:A Behavior Analysis Approach to Computer Systems Analysis
Area: EDC; Domain: Theory
CELSO S. OLIVEIRA (University of Sao Paulo)
Abstract: For years System Analysis has been an important tool to build quality software in Computer Science. Since the creation of UML - Unified Modelling Language, that uses 13 diagrams to represent the basic architecture of computer systems and its performance, the development of such systems has changed from structural to object approach. Although the Object Oriented Analysis has turned to be the main approach taught around the word, this approach doesn't quite explain how an object interacts within it's universe. An experiment of teaching basic classic Behavior Analysis concepts to Computer Science Undergraduate students of three semesters of Software Engineering discipline has showed that those concepts fits into the diagrams to explain the behavior of computer systems, how the people and other devices relates and behaves while still in the development phase of the software and improved the students skills of designing better software.
39. Ex-Posto Facto Analysis of Influence of Feedback on Enhancing Students' Learning in University Class
Area: EDC; Domain: Applied Research
MASAKO YOSHIOKA (Aichi University)
Abstract: A teacher and students strongly influence each other. Investigating their activity is important to find feasible variables to enhance spontaneous learning. For this purpose, this study conducted ex-post facto analysis of influence of teachers feedback on undergraduates written report in two classes. Students wrote their questions and thoughts in communication-sheet about the given lecture. The written reports were classified into following five categories, question, answer to the quiz, difficulty, teaching method, and others. Main target of the intervention was to increase reports of question, because they are supposed to enhance creative learning. The intervention was consisted of 1) showing a graph of the distribution of reports by category (i.e. question" etc.), 2) reply to questions and some thoughts, and 3) praise for the target reports. In both classes, the number of students who reported question and the number of letters in the reports increased during the first several sessions. Also, some students showed positive reports to feedback. Then the data showed, however, a gradual decrease. It is considered that the limited effects could be improved by motivational operation related to creative thinking behavior.
40. Charting the Future Course of SAFMEDS
Area: EDC; Domain: Theory
SCOTT A. MILLER (University of Nebraska Medical Center/Brohavior), Michael Fantetti (Western New England University/Brohavior), Cameron Green (Florida State University, Brohavior  ), Ryan Lee O'Donnell (Brohavior), Mark Malady (Brohavior; HSI/WARC)
Abstract: A common form of instructional delivery employing Lindsley's (1996) four free-operant freedoms is a flashcard technique called Say All Fast a Minute Each Day Shuffled, or "SAFMEDS" (Lindsley, 1984). However, there are currently several disparate recommended methods for learning with SAFMEDS. Some examples are separated instruction, pre-instruction, or feedback within the timed practice session, as well as others. To date, there have been relatively few direct comparisons of these components with respect to celeration rates for see/say facts. The current investigation compared fluency timings alone to practice-then-fluency timings on the rate of celeration and maintenance for see/say multiple math facts. Results from one college-aged participant demonstrated higher levels of accuracy and celeration values for flashcards in the practice-then-fluency group. The alternative fluency-only group demonstrated lower celebration values and took longer to achieve the mastery criteria (the fluency aim). Implications suggest exposure to all relevant stimuli may be the requisite mechanism responsible for skill acquisition. Future directions for research are discussed.
41. Evaluating Behavioral Research Skills of Students in a Psichology Master Program.
Area: EDC; Domain: Applied Research
MARCO WILFREDO SALAS-MARTINEZ (University of Veracruz, Mexico), Hilda Lopez Dominguez (University of Veracruz, Mexico), Esperanza Ferrant-Jimenez (University of Veracruz), Martin Ortiz Beno (University of Veracruz, Mexico), Minerva Perez Juarez (University of Veracruz, Mexico)
Abstract: The low rates of students obtaining their degree is a problem at the National and state level. In Veracruz, México, only a forty percent of students complete their master degree. The Master Applied Research in Educational Psychology Program reports a low graduation rate, above the one required by the University of Veracruz and the National Council of Science and Technology. The purpose of this study was to evaluate the effect of a Thesis Supervision System (Malott and Salas, 2010) on the development of behavioral research skills of eleven students of the master program. A Within Subjects Design (Hersen and Barlow 1977) was used to record and evaluate the research tasks of the eleven students in each phase: planning or writing a proposal, implementation or intervention and report writing. It was found that the Thesis Supervision System is efficient for the evaluation and increasing the behavioral research skills, since all students successfully completed the tree phases, regardless of their experience on research, training and other factors.
42. A Digital Method for Developing Reading and Writing Skills for Illiterate Adults
Area: EDC; Domain: Applied Research
MARCO WILFREDO SALAS-MARTINEZ (University of Veracruz, Mexico), Esperanza Ferrant-Jimenez (University of Veracruz), Amelia Sarco Santo (University of Veracruz, Mexico), Christian Mendoza Juarez (University of Veracruz, Mexico), Rafael de Jesus Jacome Serena (University of Veracruz, Mexico), Martin Ortiz Beno (University of Veracruz, Mexico), Graciela Patricia Huerta Giles (University of Veracruz, Mexico)
Abstract: Illiteracy is one of the most serious socio- educational problems facing by Mexican government Veracruz ranks fourth nationally with an average of 619.394 illiterate adults. According to the National Institute of Statistics, Geography and Informatics (INEGI, 2010). Therefore, the purpose of the study was to evaluate the effect of a program for developing digital literacy skills in a sample of adults illiterate in Xalapa, Veracruz. Digital Literacy Program Writing and Reading for Adults (PADELA) also enables basics skills for managing computer. It is a program based on programmed instruction with which adults heard and saw the instructions, getting feedback on their exercises, and learning at their own pace. Participated in this study five illiterate female adults aged 35-70 years. RAVEN test was used to measure intellectual ability of adults, also pre-and post-test assessments of skills development in each verbal unit program were applied. For the study, a Signal Case Design (Ary, Jacobs and Razavieh, 1987) was used. The results demonstrate that the method using digital programmed instruction was effective to develop reading and writing skills in the five adults.
43. Comparing Fluency and Accuracy: Response Rates at a Four-Week Follow-Up
Area: EDC; Domain: Theory
SCOTT A. MILLER (University of Nebraska Medical Center), Cameron Green (Florida State University, Brohavior), David A. Wilder (Florida Institute of Technology)
Abstract: Views in teaching have provided opposing arguments regarding measurement and mastery criteria related to rate of response (fluency). The purpose of this research was to extend literature comparing non-time based performance criteria against time-based performance criteria with regard to maintenance after four weeks without instruction. We evaluated the frequency and accuracy of labeling foreign alphabet characters four weeks after training using either an accuracy-only criterion or a fluency-based criterion. The two conditions were compared for their respective effects on maintenance and preference. Two undergraduate psychology students were taught to identify alphabet characters from foreign languages arbitrarily corresponding with English letters from two sets of 80 cards. Teaching continued until each participant reached her terminal performance criteria for each set of cards. Results showed that achieving fluency aims required a greater number of trials than achieving accuracy. Further, results indicated that the performance maintained at slightly higher levels for cards trained in the accuracy condition. Implications and future research directions are discussed.
44. Elementary Statistics Courses Fail Miserably in Teaching the P-Value
Area: EDC; Domain: Basic Research
Bradley E. Huitema (Western Michigan University), JESSICA L. URSCHEL (Western Michigan University)
Abstract: A key concept that is taught in almost every elementary statistics course is the p-value. Students in both behavioral and traditional psychology undergraduate programs are expected to know what this value means. Although the usefulness of the p-value has been widely debated, the majority of peer-reviewed psychology journals outside the behavioral arena expect either confidence intervals or p-values to be reported. The purpose of this study was to evaluate whether students who had taken at least one undergraduate statistics course could correctly interpret a p-value. A sample (N = 391) of students (from various departments and universities) who had taken an undergraduate statistics course was surveyed at the beginning of a graduate-level statistics course. Each student was asked to select all correct interpretations of a given p-value from a list of six options. Only 28 students (0.7%) correctly identified the only correct interpretation listed. The majority of respondents confused statistical significance with evidence of an important effect. It is concluded that elementary statistics courses generally fail in conveying the meaning of p-values.
Keyword(s): poster session



Back to Top
Modifed by Eddie Soh