The Paradox of Choice in fictitious COVID-19 vaccination scenario: the role of the number of options and the amount of information in decision-making.
Dublin Core
Title
The Paradox of Choice in fictitious COVID-19 vaccination scenario: the role of the number of options and the amount of information in decision-making.
Creator
Iveta Volna
Date
2021
Description
Previous research evidence showed that when people face abundance of choices or too much information, they tend to experience the paradox of choice. This study investigates the role of the number of options and amount of information in decision-making, respectively, the paradox of choice in the fictitious COVID-19 vaccination scenario. Participants (N = 128) were randomly allocated to one of four experimental conditions. The conditions differed in the number of options (high – six options; low – two options) and the information (high – six pieces of information per option; low – two pieces of information per option). As a result, the four experimental conditions were: low options, low information; low options, high information; high options, low information; high options, high information. Participants were asked to choose one of the vaccines from a list presented separately from the experimental stimuli. The reaction time of choosing a vaccine was measured. Participants were asked to evaluate how satisfied they were with their choice, how confident they were about their choice and their anticipated regret. Participants were also asked to write the reason why they chose a particular option. The study did not find a significant effect of the number of options and the amount of information on the decision-making. Participants identified five main themes why they chose a particular option: features of the vaccine, scientific evidence, information, lawfulness, and personal preference. The study revealed positive relationships between choice satisfaction, confidence, and anticipated regret.
Source
Participants
The participants’ pool was collected from the general public. In total, 191 participants took part in the study. However, 62 participants were excluded due to missing values. Another participant was further excluded because of stating being 0 years of age. Thus, the data of 128 participants (36 males, 90 females, 1 non-binary/third gender, and 1 prefer not to say) were used for the analysis. The participants were in the age group between 18 and 51 years of age (M = 23.1, SD = 6.02). Based on power analysis when effect size f = .25 (medium effect), p = .05, power (1 – β error probability) = .80, and the number of groups = 4, it was indicated that the sample size of 128 participants is necessary to ensure the study results have high statistical power. In terms of age, one participant was stated to be 22,5 years old. For the analysis, this was taken as 22 years of age. The participants were invited to the research via Facebook post, Instagram story, and direct messaging friends and family circles.
From the overall sample, 103 participants (29 males, 72 females, 1 non-binary/third gender, 1 prefer not to say) also filled an additional qualitative question investigating the reasoning behind the participants’ choice. As drawn from the overall sample, the age of participants responding to the qualitative question ranged from 18 to 47 years of age (M = 22, SD = 5.76).
Design
Participants were presented with information about fictitious COVID-19 vaccines. The current study applied a 2x2 between-subject design. Participants were randomly split according to the number of options (high – six options; low – two options) and the amount of information they received (high – six pieces of information per option; low – two pieces of information per option). Consequently, the four experimental conditions were:
a) Low options, low information
b) Low options, high information
c) High options, low information
d) High options, high information
The six vaccines in the high options scenario represented the first six COVID-19 vaccines used in the world in more than two countries (Forbes, 2021). Two vaccines in the low option scenario were chosen as it is the smallest number of options participants can compare and choose from. The amount of information then copied the design of the number of options. The number of options and the amount of information was then counterbalanced, enabling testing the effect of the number of options versus the amount of information and their interaction on decision-making.
Materials
As mentioned above, the data was collected using an online questionnaire. Participants were randomly allocated to one of the experimental conditions using the Qualtrics.com question randomiser function. Thus, there was no control of the researcher regarding the experimental condition allocation.
The experimental stimuli consisted of pictures containing the information about vaccines varying in the number of information and the number of vaccines, as can be seen in Figure 1 to Figure 4. The information about each vaccine in the experimental stimuli was inspired by the real-world COVID-19 vaccines in use. For collecting the information, official sources were reviewed, news articles and videos, and other websites. Although the information was modified, it does not directly correspond with any real-world vaccine. The sources also do not directly match with real-world sources of information. All people, social media accounts, and websites are fictitious. The information was counterbalanced, so each of the fictitious vaccines has a similar amount of information from official sources (CDC, NHS, WHO, Government) and unofficial sources (made up websites and social media profiles). Further, to ensure there is no dominant option, the number of people in fictitious vaccine trials was similar. Likewise, the efficiency levels were kept similar across the options, and positive and negative information was also balanced.
Figure 1
Experimental stimulus – Low information, low options
Note. This picture represents the low information, low options experimental condition presenting two pieces of information and two vaccines options.
Figure 2
Experimental stimulus – Low information, low options
Note. In this picture, the low option high information experimental condition can be seen. Two options of vaccines and six pieces of information for each vaccine were presented in this condition.
Figure 3
Experimental stimulus – Low information, high options
Note. This figure illustrates the high options and low information experimental condition. In this experimental condition, six vaccines and two pieces of information for each vaccine were presented.
Figure 4
Experimental stimulus – High information, high options
Note. The high options, high information condition stimulus was split into two pictures to assure that the font of the text is sufficiently large for the participants to read the information. Six options of vaccines and six pieces of information to each vaccine were presented in this condition.
After viewing the stimuli, participants were asked to select one vaccine from the list on a separate page. The lists of vaccines varied depending on the experimental conditions—the number and type of vaccines corresponded with the experimental stimulus. The page with vaccines options was timed to measure how long participants spend deciding between the vaccines. The time was measured from when the page came up until submitting the page. For the open-ended question about the rationale behind the choice, a larger text box was provided so the participants could type in a short paragraph about why they decided on that vaccine. 5-point Likert scales were used to measure satisfaction (unsatisfied to satisfied), confidence (unconfident to confident), and regret (regret to not regret at all).
Procedure
In the beginning, participants were informed about the nature of the experimental task; however, they were not told that the study measures the paradox of choice. Participants could continue the study after completing a consent form.
Then, participants were informed that they would view lists of information about fictitious COVID-19 vaccines. They were recommended to take notes to maximise their attention to the information. Then participants proceeded to one of the experimental conditions and were asked to read through the information presented. Then they continued to another page and were asked to choose one of the vaccines from the list based on the information from the previous page. The questionnaire continued with the open-ended question. The following page contained the evaluation of the choice satisfaction, confidence, and regret. Participants were disclosed that the paradox of choice was measured in the debriefing, followed by its definition and links to the actual COVID-19 vaccines information. The participants were given the option to withdraw by closing the browser window without saving their data if they no longer wished to participate in the study. The experimental design was reviewed and approved by the Lancaster University Department of Psychology ethical committee.
Data analysis methods
This study investigates the effect of the number of options and the amount of information on the paradox of choice across the four experimental conditions. The dependent variables measured were the reaction time, satisfaction levels, choice confidence, and anticipated regret measured using a 5-point Likert scale. The data gathered consists of independent observations as everyone went through one experimental condition at the time. Convenient sampling was used to collect data as the participants were mainly the researcher’s family, friends, and acquaintances. However, the participants come from different countries, age categories and educational backgrounds; thus, it can be assumed that the observations are independent of each other. The effect of two factors (information, options) with two levels (low and high) on dependent variables are observed. Hence, a 2x2 analysis of variance (ANOVA) was chosen as an appropriate analysis for testing the research hypotheses.
The relationships between choice satisfaction, confidence, and anticipated regret were also investigated. The data were checked for the assumption of linearity. The data on satisfaction seems to be positively skewed similarly to the data on confidence and regret. However, the data appear not to be linear nor homoscedastic. Therefore, Spearman’s correlation was chosen as an adequate analysis for this type of data.
The short responses to the qualitative question “Why did you decide on that option?” was analysed using template analysis (King & Brooks, 2017). Template analysis is a flexible type of thematic analysis that can be used to analyse written responses to an open-ended question on a questionnaire (Brooks et al., 2015; King & Brooks, 2017). The question about the participants’ rationale behind their choice used in the current study was open-ended. Participants were asked to give a short written answer. Because the data result from an open-ended question and the flexibility of template analysis, template analysis was chosen to analyse the quantitative data. The final template is presented below in Figure 5 in the results section. In the beginning, all data was put together in one document. Next, the participants’ answers were coded line by line. Then the line coding was used to identify themes. The second level of themes was generated from the first level themes. Because the answers consist of short sentences with a maximum of short paragraphs, two levels of themes were used in the analysis. The template was developed from the two levels of codes. The template checked whether it fits all the recorded answers ensuring the template accuracy. Then the template was reviewed and concluded, referring to the sixth research question.
The participants’ pool was collected from the general public. In total, 191 participants took part in the study. However, 62 participants were excluded due to missing values. Another participant was further excluded because of stating being 0 years of age. Thus, the data of 128 participants (36 males, 90 females, 1 non-binary/third gender, and 1 prefer not to say) were used for the analysis. The participants were in the age group between 18 and 51 years of age (M = 23.1, SD = 6.02). Based on power analysis when effect size f = .25 (medium effect), p = .05, power (1 – β error probability) = .80, and the number of groups = 4, it was indicated that the sample size of 128 participants is necessary to ensure the study results have high statistical power. In terms of age, one participant was stated to be 22,5 years old. For the analysis, this was taken as 22 years of age. The participants were invited to the research via Facebook post, Instagram story, and direct messaging friends and family circles.
From the overall sample, 103 participants (29 males, 72 females, 1 non-binary/third gender, 1 prefer not to say) also filled an additional qualitative question investigating the reasoning behind the participants’ choice. As drawn from the overall sample, the age of participants responding to the qualitative question ranged from 18 to 47 years of age (M = 22, SD = 5.76).
Design
Participants were presented with information about fictitious COVID-19 vaccines. The current study applied a 2x2 between-subject design. Participants were randomly split according to the number of options (high – six options; low – two options) and the amount of information they received (high – six pieces of information per option; low – two pieces of information per option). Consequently, the four experimental conditions were:
a) Low options, low information
b) Low options, high information
c) High options, low information
d) High options, high information
The six vaccines in the high options scenario represented the first six COVID-19 vaccines used in the world in more than two countries (Forbes, 2021). Two vaccines in the low option scenario were chosen as it is the smallest number of options participants can compare and choose from. The amount of information then copied the design of the number of options. The number of options and the amount of information was then counterbalanced, enabling testing the effect of the number of options versus the amount of information and their interaction on decision-making.
Materials
As mentioned above, the data was collected using an online questionnaire. Participants were randomly allocated to one of the experimental conditions using the Qualtrics.com question randomiser function. Thus, there was no control of the researcher regarding the experimental condition allocation.
The experimental stimuli consisted of pictures containing the information about vaccines varying in the number of information and the number of vaccines, as can be seen in Figure 1 to Figure 4. The information about each vaccine in the experimental stimuli was inspired by the real-world COVID-19 vaccines in use. For collecting the information, official sources were reviewed, news articles and videos, and other websites. Although the information was modified, it does not directly correspond with any real-world vaccine. The sources also do not directly match with real-world sources of information. All people, social media accounts, and websites are fictitious. The information was counterbalanced, so each of the fictitious vaccines has a similar amount of information from official sources (CDC, NHS, WHO, Government) and unofficial sources (made up websites and social media profiles). Further, to ensure there is no dominant option, the number of people in fictitious vaccine trials was similar. Likewise, the efficiency levels were kept similar across the options, and positive and negative information was also balanced.
Figure 1
Experimental stimulus – Low information, low options
Note. This picture represents the low information, low options experimental condition presenting two pieces of information and two vaccines options.
Figure 2
Experimental stimulus – Low information, low options
Note. In this picture, the low option high information experimental condition can be seen. Two options of vaccines and six pieces of information for each vaccine were presented in this condition.
Figure 3
Experimental stimulus – Low information, high options
Note. This figure illustrates the high options and low information experimental condition. In this experimental condition, six vaccines and two pieces of information for each vaccine were presented.
Figure 4
Experimental stimulus – High information, high options
Note. The high options, high information condition stimulus was split into two pictures to assure that the font of the text is sufficiently large for the participants to read the information. Six options of vaccines and six pieces of information to each vaccine were presented in this condition.
After viewing the stimuli, participants were asked to select one vaccine from the list on a separate page. The lists of vaccines varied depending on the experimental conditions—the number and type of vaccines corresponded with the experimental stimulus. The page with vaccines options was timed to measure how long participants spend deciding between the vaccines. The time was measured from when the page came up until submitting the page. For the open-ended question about the rationale behind the choice, a larger text box was provided so the participants could type in a short paragraph about why they decided on that vaccine. 5-point Likert scales were used to measure satisfaction (unsatisfied to satisfied), confidence (unconfident to confident), and regret (regret to not regret at all).
Procedure
In the beginning, participants were informed about the nature of the experimental task; however, they were not told that the study measures the paradox of choice. Participants could continue the study after completing a consent form.
Then, participants were informed that they would view lists of information about fictitious COVID-19 vaccines. They were recommended to take notes to maximise their attention to the information. Then participants proceeded to one of the experimental conditions and were asked to read through the information presented. Then they continued to another page and were asked to choose one of the vaccines from the list based on the information from the previous page. The questionnaire continued with the open-ended question. The following page contained the evaluation of the choice satisfaction, confidence, and regret. Participants were disclosed that the paradox of choice was measured in the debriefing, followed by its definition and links to the actual COVID-19 vaccines information. The participants were given the option to withdraw by closing the browser window without saving their data if they no longer wished to participate in the study. The experimental design was reviewed and approved by the Lancaster University Department of Psychology ethical committee.
Data analysis methods
This study investigates the effect of the number of options and the amount of information on the paradox of choice across the four experimental conditions. The dependent variables measured were the reaction time, satisfaction levels, choice confidence, and anticipated regret measured using a 5-point Likert scale. The data gathered consists of independent observations as everyone went through one experimental condition at the time. Convenient sampling was used to collect data as the participants were mainly the researcher’s family, friends, and acquaintances. However, the participants come from different countries, age categories and educational backgrounds; thus, it can be assumed that the observations are independent of each other. The effect of two factors (information, options) with two levels (low and high) on dependent variables are observed. Hence, a 2x2 analysis of variance (ANOVA) was chosen as an appropriate analysis for testing the research hypotheses.
The relationships between choice satisfaction, confidence, and anticipated regret were also investigated. The data were checked for the assumption of linearity. The data on satisfaction seems to be positively skewed similarly to the data on confidence and regret. However, the data appear not to be linear nor homoscedastic. Therefore, Spearman’s correlation was chosen as an adequate analysis for this type of data.
The short responses to the qualitative question “Why did you decide on that option?” was analysed using template analysis (King & Brooks, 2017). Template analysis is a flexible type of thematic analysis that can be used to analyse written responses to an open-ended question on a questionnaire (Brooks et al., 2015; King & Brooks, 2017). The question about the participants’ rationale behind their choice used in the current study was open-ended. Participants were asked to give a short written answer. Because the data result from an open-ended question and the flexibility of template analysis, template analysis was chosen to analyse the quantitative data. The final template is presented below in Figure 5 in the results section. In the beginning, all data was put together in one document. Next, the participants’ answers were coded line by line. Then the line coding was used to identify themes. The second level of themes was generated from the first level themes. Because the answers consist of short sentences with a maximum of short paragraphs, two levels of themes were used in the analysis. The template was developed from the two levels of codes. The template checked whether it fits all the recorded answers ensuring the template accuracy. Then the template was reviewed and concluded, referring to the sixth research question.
Publisher
Lancaster University
Format
data/r.csv
Identifier
Volna2021
Contributor
Faye Summers
Connie Jordan-Turner
Connie Jordan-Turner
Rights
Open (unless stated otherwise)
Relation
None (unless stated otherwise)
Language
English
Type
Data
Coverage
LA1 4YF
Files
Collection
Citation
Iveta Volna, “The Paradox of Choice in fictitious COVID-19 vaccination scenario: the role of the number of options and the amount of information in decision-making.,” LUSTRE, accessed May 16, 2024, https://www.johnntowse.com/LUSTRE/items/show/121.