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.

Search

47th Annual Convention; Online; 2021

All times listed are Eastern time (GMT-4 at the time of the convention in May).

Event Details

Previous Page

 

Poster Session #87
Saturday, May 29, 2021
1:00 PM–3:00 PM
Online
6.

Effects of Differential Reinforcer Magnitude of an Alternative Response on the Resurgence of Academic Responding

Area: EAB; Domain: Applied Research
EMILY L. BAXTER (Syracuse University), Brian K. Martens (Syracuse University), Taysha Cerisier (Syracuse University), Samantha Sallade (Syracuse University), Joshua Circe (Syracuse University)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract:

Several studies have looked at ways to mitigate resurgence of a target behavior by manipulating dimensions of reinforcement for an alternative behavior. To date, only one study has examined differences in resurgence following different magnitudes of reinforcement for the alternative behavior, and only one study has addressed resurgence in an academic setting. The current study evaluated resurgence of a target academic response when all responses were placed on extinction subsequent to a phase of high- or low-magnitude reinforcement for an alternative response. Four fourth-grade students without a diagnosis participated and their rate of problem completion was measured across sessions. In Phase 1, students were reinforced for completing addition problems. In Phase 2, students were reinforced for completing subtraction problems, but not addition problems. Finally in Phase 3A/3B, reinforcement was not provided for any response (i.e., extinction). In Phase 3A, extinction was not signaled and resurgence was only observed in 3 of the 4 students. In Phase 3B, extinction was signaled and resurgence of the target response occurred for all four participants, with variable levels across the high- and low-magnitude conditions. Implications and directions for future research will be discussed.

 
8. Cross-Cultural Comparison of Delay Discounting in American and Japanese College Students
Area: EAB; Domain: Basic Research
MICHIKO SORAMA (Kyoto Notre Dame University), Lori-Ann B. Forzano (The College at Brockport, State University of New York), Michael Fensken (State University of New York-Brockport), Cara Bakalik (State University of New York-Brockport), Lauren Soda (State University of New York-Brockport), Lauren Teti (State University of New York-Brockport), Heather Graupman (State University of New York-Brockport)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract: Impulsivity is fundamental in many unhealthy behaviors and is also featured in several psychological disorder diagnoses including, Attention-Deficit/Hyperactivity Disorder and Substance Abuse. Delay discounting is the process by which one diminishes or devalues delayed rewards, and this process underlies some forms of impulsivity. Because delay discounting plays a role in maladaptive behaviors, understanding additional factors related to it, is essential. Previous research has examined the role of culture in relationship to delay discounting. Specifically, it has been found, among adults, Americans are more likely than Japanese to discount future reinforcers. However, because other studies have found age differences in delay discounting, the current study, compared delay discounting across cultures in adolescents and emerging adults, i.e., college students, who completed the Delay Discounting Questionnaire (Sorama et al., 2019). Analysis of log k of 234 American (M = -1.296, SD = 0.74) and 91 Japanese (M = -1.325, SD = 0.76) participants revealed no significant difference in delay discounting, t(323) = 0.3199, p = .7492. This study suggests other factors that may underlie cross-cultural comparisons in delay discounting.
 
9.

The Merger of Equivalence Classes Established Experimentally and Pre-Experimentally via a Common Stimulus (Printed Word)

Area: EAB; Domain: Basic Research
RAMON MARIN (Universidade Federal de São Carlos ), Vanessa Pereira (University of Bergen), Deisy De Souza (Universidade Federal de São Carlos)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract:

Meaningful stimuli are, by definition, pre-experimentally related to other stimuli. Some authors have suggested that training relations between a meaningful stimulus and a member of an equivalence class could produce the merge of the experimental class with a pre-experimentally established class. Here, two experiments compared the effects of meaningful and meaningless words on the merger of classes. In both experiments, adults were trained on AB and AC relations for the emergence of three 3-member equivalence classes with abstract stimuli. Then, they were trained on DA relations and tested for the inclusion of D and a set of nine meaning-related pictures in the class. In Experiment 1 (N = 22), D1 and D2 were meaningful written words, and D3 was a written pseudo word. In Experiment 2 (N = 13), for two groups, D1, D2, and D3 were meaningful words; for another group, D1, D2, and D3 were meaningless words. In both experiments, only meaningful words produced the merger of the ABC and the meaning-related stimulus classes. These results support the notion that meaningful stimuli are members of pre-experimental classes and that using meaningful-stimulus in equivalence tasks can produce the merge of the experimental and the extra-experimental classes.

 
10. Further Analysis of Mixed-Compound Consequences and Their Role in Equivalence-Class Formation
Area: EAB; Domain: Basic Research
RICHELLE ELIZABETH HURTADO (University of North Carolina Wilmington), Carol Pilgrim (University of North Carolina Wilmington)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract: This study investigates the impact of mixed-compound class-specific consequences on equivalence-class formation across age groups. Three conditions compare A and B simple-discrimination training with completely class-specific compound consequences (i.e., A/B1→R1r1, A/B2→R2r2, A/B3→R3r3); A and B discrimination training with mixed compound consequences, in which one element is class-specific and one is common across contingencies (i.e., A/B4→R0r4, A/B5→R0r5, A/B6→R0r6); and A and B discrimination training with a common compound consequence for all discriminations (i.e., A/B7→R0R0, A/B8→R0R0, A/B9→R0R0). Conditional discrimination probes assess emergent relations between A, B, R, and r stimuli. The data from 6- to 8-year-old children indicate that equivalence-class formation was impeded by initial exposure to the mixed-compound-consequence condition. Two of three children trained first with the mixed-compound consequences demonstrated emergent relations between the reinforcer elements and the discriminative stimuli (i.e., A and B). However, none of the three participants demonstrated emergent AB relations (see Table 1), in contrast to the children trained initially with non-mixed compound stimuli. Data from college students are being collected currently. Similar patterns of dissociation in emergent performances across developmental groups would strengthen the case that mixed-compound consequences hinder class formation and would support Sidman’s (2000) theory that equivalence is a direct outcome of reinforcement contingencies.
 
11. Reorganization and Maintenance of Equivalence Classes after Overtraining
Area: EAB; Domain: Basic Research
GIOVAN WILLIAN RIBEIRO (Universidade Federal de São Carlos), Deisy De Souza (Universidade Federal de São Carlos)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract: Stimulus equivalence is an important phenomenon to the understanding of symbolic behavior. We investigated the maintenance of equivalence classes after overtraining either baseline or reversed relations. Participants’ accuracy and responding speed were measured. Nineteen undergraduates performed a three-choice match-to-sample to learn baseline conditional discriminations AB, AC, and AD. These relations were then mixed in a review block. The Formation Overtraining Group (FOt; N=9) performed 81 additional review trials. Equivalence tests showed the formation of three classes: A1B1C1D1, A2B2C2D2, and A3B3C3D3. Following the tests, AD trials were reversed (A1D2, A2D3, A3D1) and a new review block conducted, now containing the baseline relations AB and AC and AD reversed. The Reorganization Overtraining Group (ROt; N=10) performed 81 additional review trials. Reorganization tests showed that the classes reorganized into A1B1C1D2, A2B2C2D3, and A3B3C3D1. The ROt group responded significantly faster than FOt in these tests. After 14-30 days, the class maintenance was tested. Four participants from the FOt group maintained the baseline classes, while one from each group maintained the reorganized classes. The FOt group responded faster than the ROt group in the maintenance trials. These results indicate that overtraining enhances class reorganization, but reorganized classes are hardly maintained over time.
 
12.

Using Probability and Social Discounting to Predict Compliance With Protective Measures During the COVID-19 Pandemic

Area: EAB; Domain: Basic Research
JULIO CAMARGO (Universidade Federal de São Carlos), Denise Aparecida Passarelli (Universidade Federal de São Carlos), Marlon Alexandre de Oliveira (Universidade Federal de São Carlos), Filipe César Carvalho (Universidade Federal de São Carlos), Djenane Brasil da Conceição (Universidade Federal do Recôncavo da Bahia), Josiane Maria Donadeli (Universidade Federal de São Carlos), Alceu Regaço (Universidade Federal de São Carlos), Julio C. De Rose (Universidade Federal de São Carlos)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract:

The purpose of the present study was to investigate if the behavioral processes of probability and social discounting could predict people’s compliance with the protective measures recommended during the COVID-19 pandemic. A sample of 116 adults living in Brazil completed an online survey composed of a probability discounting questionnaire (PDQ), a social discounting questionnaire (SDQ), and a 10-question assessment of how often they complied with the health authorities’ recommendations (e.g., wash the hands frequently, practice social distancing, stay at home as much as possible, wear a mask when in public). Only the participants who showed higher or lowered compliance with the protective measures were included in the final analyses. Results revealed that the PDQ and SDQ responses could predict people’s compliance with the protective measures recommended by the health authorities. Participants who showed low compliancewere more likely to perform risk choices in the PDQ than high-compliant ones. To a lesser extent, participants who showed high compliance were more prone to present cooperative choices in the SDQ, compared to low-compliant participants. These results suggest that probability and social discounting processes can play an essential role in the self-protective and cooperative decisions people made during a global health emergency, such as the COVID-19 pandemic.

 
13.

The Function of Observing Responses: Investigating the Effects of Observing-Stimulus Duration

Area: EAB; Domain: Basic Research
PETER KIM (University of Auckland), Douglas Elliffe (University of Auckland), Sarah Cowie (University of Auckland, New Zealand)
Discussant: Erik Arntzen (Oslo Metropolitan University)
Abstract:

Six pigeons pecked on the food key to receive reinforcers on a two-component mixed schedule (VI 30 s/EXT), or on the observing key to change the mixed schedule to a multiple schedule by sometimes producing an observing stimulus (S+/S-) that corresponded to the current component in effect. Components alternated irregularly with variable durations and responses within the final 5 s of an EXT component extended the said component. Across conditions we manipulated the duration of the observing stimulus (3.25, 7.5, 15, 30 s). As observing-stimulus duration increased, percentage of session time spent as a multiple schedule increased and fewer observing stimuli were produced. What remained invariant across conditions was the degree of differential responding during the mixed stimulus. Number of successive food-key responses since offset of observing stimuli showed that observing stimuli (S+/S-) produced differential responding on the food key during the mixed stimulus. Consequently, producing observing stimuli resulted in better discrimination during the mixed stimulus compared to a mixed schedule without such an opportunity. These findings show that observing stimuli were produced to an extent that affords greater discrimination of an uncertain environment regardless of the duration of these observing stimuli.

 
14. Discounting Fixed-Ratio Requirements in Rats
Area: EAB; Domain: Basic Research
EMILY BROOKS (Central Michigan University), Emily Boley (Central Michigan University), Katie Monske (Central Michigan University), Eric James French (Central Michigan University), Mark P. Reilly (Central Michigan University)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: In concurrent schedule arrangements with a small fixed-ratio (FR) requirement on one lever and a larger FR requirement on the other lever, rats show near exclusive preference for the smaller FR. Would rats prefer the same smaller ratio when both ratios are temporally removed from the initial choice? In Phase 1 rats were exposed to a concurrent FR 5, FR 15 schedule to ensure they preferred the smaller FR schedule. In Phase 2, the procedure was modified such that FR 5 schedules were arranged on both levers, however depending on which lever was chosen, the rat had to complete another FR 5 or an FR 15 on the chosen lever. For example, if the FR 5 was completed on the right lever, food would be delivered, and another FR 5 occurred on the same lever. If the FR 5 was completed on the left lever, food would also be delivered but now an FR 15 occurred on the same lever. All rats reliably preferred the FR 5 in Phase 1. It is predicted that rats will continue to prefer the smaller FR in Phase 2 despite its temporal separation from choice and the delivery of food between consecutive ratios.
 
15.

Influencing Saliva, But Not Evaluations, Following Subliminal Conditioning

Area: EAB; Domain: Basic Research
DENISE APARECIDA PASSARELLI (Universidade Federal de São Carlos), Micah Amd (University of the South Pacific), Marlon Alexandre de Oliveira (Universidade Federal de São Carlos), Julio C. De Rose (Universidade Federal de Sao Carlos)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract:

Associating meaningless trigrams (CS) with a positively valenced attribute (US) may influence appetite-associated motivational systems, such as increased saliva production, regardless of whether the former (CS-US) relation is consciously tractable. We test this claim in the present study, where we expose three groups of Brazilian undergraduates (N = 69) to a subliminal conditioning protocol. The procedure involved presenting eating-related activities (CS) or their scrambled counterparts (non-CS) at subliminal visual thresholds, followed by visible positive or neutral attributes (US). A free-selection visibility check confirmed that no participant identified the subliminal stimuli during the conditioning task. For the group who associated eating activities with the positive US, saliva production increased significantly as predicted. Outcomes across explicit and implicit evaluations were statistically inconclusive, but Markov Chain Monte Carlo (MCMC) simulations suggested conditioning effects were somewhat likely. Reliable saliva augmentation, coupled with inconclusive evaluation outcomes, illustrates how affective CS-US information can influence motivational responses with minimal deliberative influence.

 
16. Social Discounting Towards Relatives and Non-relatives
Area: EAB; Domain: Basic Research
NATALIE BUDDIGA (University of Nevada, Reno), Matt Locey (University of Nevada, Reno)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: Social discount functions have been used as a quantitative description of altruistic choice. In support of kin altruism theories, previous research has found that participants discounted outcomes less to relatives than non-relatives. However, weaknesses in those analyses, such as disproportionate value comparisons, limit the validity of those findings. The current study sought to amend those weaknesses by using individual social discount rates and explicit identification of the beneficiary to assess differences in altruistic choice. Results demonstrated the same direction of altruism in favor of relatives like previous research, but a multilevel analysis demonstrated no significant difference between discounting towards relationship types. Moreover, explicit identification of beneficiaries did not alter altruism significantly. Implications for future research and determinants of altruistic choice are discussed.
 
17. Human Conc FR FI Responding in a Computer Game May Shows Sensitivity to Schedule Contingencies
Area: EAB; Domain: Basic Research
DEBRA J. SPEAR (South Dakota State University), Eleanor Dick (South Dakota State University)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: In similar experimental settings, humans do not always respond like rats and pigeons. It is not as straight forward to determine if they are sensitive to schedule contingencies as it is for rats and pigeons. In two studies, college students competed a 3-component Mult Conc FR FI designed as a computer game. The schedule in the first study was Mult Conc FR 20 FI 30; FR 100 FI 30; FR 150 FI 30. The schedule in the second study was Mult Conc FR 100 FI 10; FR 100 FI 20; FR 100 FI 30. Response rates were typically higher in the FR component, regardless of the value of the Conc FI value and regardless of a higher number of reinforcers following FI responding. Despite the lack of obvious sensitivity to the contingency and a strong preference for the FR schedule, there was evidence that schedule values impacted responding. However, this sensitivity is not consistent for all participants and may be related to self-generated rules.
 
18. Effects of an Intruding Stimulus on the Temporal Distribution of Schedule-Induced Ethanol Consumption in Rats
Area: EAB; Domain: Basic Research
RAUL AVILA (National Autonomous University of Mexico), GUADALUPE MOGUEL (National Autonomous University of Mexico)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: A common effect in schedule-induced drinking (SID) procedures with restricted opportunity for drinking is that water consumption decreases as water is presented closer to the beginning of the inter-food intervals. To test the generality of this finding, food-deprived rats were exposed to a SID procedure with ethanol continuously available. Specifically, rats pressed a lever for food pellets in a 64-s fixed-interval (FI) schedule and pressed another one in a continuous reinforcement schedule to obtain a drop of an 8% ethanol solution. An 8-s neutral stimulus was presented within different temporal locations of the FI schedule; for four rats, it was presented from the beginning to the end of the interval, and for other four, from the end to the beginning of the interval. Adding and varying the temporal location of the stimulus modulated the temporal distribution of lever-pressing for ethanol compared to baseline conditions, as reported in previous studies with water. These findings are discussed in terms of stimulus control of timing behavior in a schedule-induced behavior paradigm.
 
19.

Functional Classes in Rats

Area: EAB; Domain: Basic Research
KYNDRA LAWSON (University of North Carolina Wilmington), Sarah Elizabeth Accattato (University of North Carolina Wilmington), Melissa Meglin (University of North Carolina Wilmington), Spencer Bruce (University of North Carolina Wilmington), Cassondra Giarrusso (University of North Carolina Wilmington), Sophie Pinneke (University of North Carolina Wilmington), Elijah Richardson (University of North Carolina Wilmington), Mark Galizio (University of North Carolina Wilmington)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract:

The present study was designed to determine whether the repeated reversal procedure that has been used to demonstrate the formation of functional classes in pigeons and sea lions could be adapted for rats using olfactory stimuli. Rats were exposed to simple discrimination training with responding reinforced on an FI 5-s schedule for six odors (Set 1 positive) and extinction for a different set of six odors (Set 2 negative). When responding was well differentiated, the discriminations were reversed (i.e., Set 2 positive, Set 1 negative). Discrimination reversals were repeatedly performed and performances were on the initial session of a reversal were analyzed to determine whether was contacting the reversed contingencies with some set members would transfer without direct training to other set members. We found transfer in most rats tested, but mainly after exposure to the changed contingencies with five of the set members. These data provide some of the first evidence of functional class formation in rats.

 
20. “Difference” Abstract Concept Learning in Rats: A Successive Nonmatching-To-Sample Procedure Using Set-Size Expansion
Area: EAB; Domain: Basic Research
MELISSA MEGLIN (University of North Carolina Wilmington), Sarah Elizabeth Accattato (University of North Carolina Wilmington), Kyndra Lawson (University of North Carolina Wilmington), Spencer Bruce (University of North Carolina Wilmington), Cassondra Giarrusso (University of North Carolina Wilmington), Sophie Pinneke (University of North Carolina Wilmington), Elijah Richardson (University of North Carolina Wilmington), Katherine Ely Bruce (University of North Carolina Wilmington)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: The current study investigates the role of multiple exemplar training with oddity relations. Using a successive, go-no-go, nonmatching-to-sample procedure, 6 Sprague-Dawley male rats were initially trained with two odorants in an automated olfactometer. A masking scent was inserted in between the presentation of the sample scent and the comparison scent. Responses on positive trials (nonmatching trials) resulted in reinforcement on a FI 5-s schedule. Once training criteria were met, eight probe sessions, each with unreinforced probe trials (four novel scents) mixed with baseline training scents, were conducted. Training set size was then expanded from two to four odorants and probe performance to four novel stimuli was assessed. Set size was then expanded to eight exemplars and probes were assessed. For two, four, and eight exemplar training, response rates were consistently higher on positive trials compared to negative trials, indicating that the rats learned the non-matching task. However, responding was typically lower and undifferentiated on probe trials when new odorants were presented. Results are compared to similar studies in rats showing transfer after four exemplars when matching-to-sample procedures were used.
 
21. Motivational State-Dependent Renewal and Reinstatement: Motivational and Discriminative Functions of Food Deprivation and Satiation Conditions
Area: EAB; Domain: Basic Research
CAITLYN PEAL (University of Nevada, Reno), Matthew Lewon (University of Nevada, Reno), Christina M. Peters (University of Nevada, Reno), Linda J. Parrott Hayes (University of Nevada, Reno)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: Recovery of extinguished responding has been observed when an organism is removed from the extinction context (renewal) and with the response-independent presentation of a reinforcer that had previously maintained the response (reinstatement). Recent studies have shown that interoceptive states associated with motivating operations (MOs) can serve as contextual stimuli that may contribute to renewal. The aim of this study was to further examine the role of deprivation/satiation states in the renewal and reinstatement of extinguished responding. In Experiments 1 and 2, four groups of mice received response acquisition and extinction sessions under four different combinations of food deprivation and satiation conditions and then received tests for renewal and reinstatement under food deprivation and satiation. State-dependent renewal and reinstatement were clearly observed under deprivation conditions but were not observed under satiation conditions, even when subjects acquired the response in this state. Experiment 3 showed that state-dependent renewal/reinstatement did not occur under satiation conditions even when mice received all pre-experimental training and response acquisition sessions under satiation conditions. The results suggest that interoceptive cues associated with MOs can contribute to the recovery of extinguished responding but that the discriminative properties of these interact with the motivational effects of MOs.
 
22. Transfer of Function and the Role of Prior Equivalence Testing
Area: EAB; Domain: Theory
HAWKEN V. HASS (College of Charleston), Adam H. Doughty (College of Charleston)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: Transfer of function typically refers to the acquisition of stimulus function by virtue of membership in an equivalence class and is a critical source of complex human behavior. Recent research re-examining the role of prior equivalence testing in generating equivalence classes suggests that there exists a complex relation. Consequently, if transfer of function is grounded in the establishment of equivalence classes, then now is a ripe time to re-examine the relation between these two concepts. Specifically, it must be determined whether, and under what conditions, derived-stimulus-relations testing is necessary to demonstrate transfer. The purpose of this project was to provide an in-depth review of the relation between prior equivalence testing and transfer of function. Nearly 60 studies involving different forms of transfer of function were reviewed. Critically, the degree to which transfer of function was obtained in these studies was analyzed as a function of the type of function transfer and multiple procedural factors. This presentation attempts to isolate the methodological variables that contribute to function transfer in the absence of equivalence testing. The implications of the review for conceptualizing transfer of function will be discussed.
 
23. Reinforcement of Variability Through An Online Video Game
Area: EAB; Domain: Basic Research
HANNAH MELELANI JOHNSON (Utah State University), Jennifer Krafft (Utah State University), Mariah Willis (Utah State University), Amy Odum (Utah State University)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: Many studies have shown that variability may be an operant dimension of behavior. In other words, antecedents and consequences can lead to highly random and unpredictable behavior. Prior research has shown that people with depression behave less variably than people without depression. The present study examined whether behavioral variability can come under reinforcement control in depressed and non-depressed participants. 70 undergraduate students were recruited from Utah State University (35 with depression and 35 without depression). Participants were given mental health measures for depression and psychological flexibility and then played an online video game that was used to measure operant variability. Consistent with previous findings, results showed that variable behavior was controlled by the reinforcement schedule. These results clearly support conclusions that variability is an operant dimension of behavior. Contrary to previous findings, however, depressed and non-depressed participants showed no difference in variable behavior under both probability and variability reinforcement schedules.
 
24. Brief Mindfulness and Human Temporal Estimation
Area: EAB; Domain: Basic Research
MARIBEL RODRIGUEZ PEREZ (California State University, Stanislaus ), Shrinidhi Subramaniam (California State University, Stanislaus)
Discussant: Paulo Morales Mayer (CEUMA)
Abstract: Previous research suggested that mindfulness meditation can increase time sensitivity and slow the passage of time. This study asked whether a brief mindfulness meditation exercise influenced temporal estimation. Participants were 115 college students who completed a verbal temporal estimation task (TE) and mindfulness exercise remotely on their personal computers or mobile devices. The TE task consisted of the random presentation of 440HZ tones with 9 durations ranging from 0.2s to 2s. Each tone was presented 5 times for a total of 45 estimates/participant. Fifty-nine of the participants were randomly assigned to complete a 5-min, audio guided, mindfulness exercise before the TE task (Mindfulness group) and the other 57 participants completed the mindfulness exercise after the TE task (Control group). We calculated the relative TE index as the difference between each temporal estimate and tone duration, divided by tone duration, for individual participants and averaged across participants in each group. Groups did not statistically significantly differ in relative TE index, but participants generally overestimated shorter durations and underestimated longer durations. Mindfulness meditation may require a controlled environment, large doses, and/or consistent practice to impact temporal estimation.
 
 

BACK TO THE TOP

 

Back to Top
Modifed by Eddie Soh
SABA DONATE