Transcription

1 Prceedings f FIKUSZ 16 Sympsium fr Yung Researchers, 2016, pp The Authr(s). Cnference Prceedings cmpilatin Obuda University Keleti Faculty f Business and Management Published by Óbuda University Under Risk Anita Klnhfer-Derecskei & Viktr Nagy derecskei.anita@kgk.uni-buda.hu; nagy.viktr@kgk.uni-buda.hu Risk is a situatin which is difficult t classify (Armenian male, 27) Abstract: The aim f this paper is t bserve hw the university students define risk. Firstly we try t explre hw ur subjects use this expressin (i.e. hw they define risk). Their answers were tested with cntent analysis technique, which helped us t highlight the mst imprtant attitudes f ur subjects. In the secnd part we fcused n five risk dmains (riginally tested by Blais & Weber, 2006). Because everybdy will have different risk attitudes, when making decisin invlving ethical, financial, health r safety, recreatinal, and scial risks. Althugh the pilt-survey was tested by MsC students this part cntained a methdlgical prblem, s this result culd nt be reliable. In the last part, accrding t the Dmain-Specific Risk Taking Scale (Blais & Weber, 2006) we try t find differences between Risk-Taking, Risk-Perceptins, and Expected Benefits. Our research shws that risk definitin culd be divided int different meaning-grups and the respndents will rder the different type f risks. Furthermre, ur results indicate hw can we use this validated psychmetric scale fr ur ppulatin in the future. Keywrds: Risk, DOSPERT Scale, Survey 1 Intrductin Risk taking is ne f the stable persnality traits which was widely studied in field f psychlgy. Ecnmists fcused n this prblem after findings f Kahneman and Tversky were published (Kahneman & Tversky, 1979) wh figured ut hw risk gives weight t ur decisins. Althugh they handled risk as a variable which definitin is bvius. The real deeper meaning, hw the subjects define it, culd be interesting as well. Thery f decisin-thery began t grw and mst f the researchers built risk-taking int his r her mdels. Hwever Weber and her clleagues (Weber, et al., 2002) suggested a validated (i.e. scientifically apprved) scale fr measurement f risk. In 2006 a new (lighted) versin was develped 161

2 which cntains nly 30 items (five risk dmains) in three different degrees. We try t rganize them int 1. Table. Dmain subscales r life dmains Risk-taking (Hw respndents engage in risky activities.) Ethical Instructin: Fr each f the fllwing Financial statements, please (Investment/Gambling) indicate the likelihd that yu Health/Safety wuld engage in the described activity r Recreatinal behavir if yu were t find yurself in Scial that situatin. 7 pints ranking scale Risk perceptin (Hw respndents asses the level f risk in each activities.) Instructin: we are interested in yur gut level assessment f hw risky each situatin r behavir is. 7 pints ranking scale Expected Benefits f risk (what kind f benefit respndents btain in each risky situatins.) Instructin: Fr each f the fllwing statements, please indicate the benefits yu wuld btain frm each situatin. 7 pints ranking scale Table 1 DOSPERT 30 (Own table based n Center fr Decisin Sciences, Clumbia Business Schl) As it culd be seen this test cntains 30 statements (all subscales have 6 statements) in three different cntexts. The authrs measured validity f test and ffered scring instructins as well (i.e. cncrete mathematical mdel hw risk can be evaluated). The test was translated int different languages, mst f all als in Hungarian. That means all subjects need t read, understand and answer the same 30 sentences cmparing 3 times, all tgether (in sum) 90 chices per a subject. Althugh it culd be handled easily with a help f IT tls but telling the truth it can be called as respndents friendly slutin. In last semester we try t use this test and it was implemented. But ur results were nt stable, unfrtunately. Maybe the surce f the prblem was that ur subjects were nt able t find differences between the befre mentined life-situatins r they were impatient t pay attentin fr all items. S the research questins was given, hw this test can be suit t ur ppulatin (in this case university students frm different cultures). But first f all hw risk culd be defined in their mind. Examinatin f the Generatin Y is increasingly imprtant, as it nt nly represents a new challenge fr the educatin system, but the labr market. (Klnhfer-Derecskei-Reicher 2016) 162

3 2 Methdlgy Regarding the literature we wrked with the items and categries frm Blais & Weber (2006). In frame f Research methdlgy curse we wrked with MsC students wh helped us t design the survey. The survey cntains fur parts and can be fund in the appendix. The first part asked the demgraphical backgrund f respndents like gender, age r natinality. Because this survey was suited nt nly fr Hungarian but freign students as well, s the survey was in English. In the secnd part respndents shuld define risk with their wn wrds. In the last half f survey we used the befre mentined scale s categries. Firstly, the answerers shuld evaluate which situatin is mre likely t happen t them. Finally, they need t judge which aspect influences their decisin. 2.1 Limitatin Despite the fact, that we tried t manage the prblem f scientifical reliability and validity. Unfrtunately, a prblem was given in the secnd part. The ppsite meaning f the instructins and scales descriptins shuld effects misunderstanding. That means, the subjects fund likelihd f the situatin and risk in the scale, which can cnfuse them. At the end we have decided that we skip ut this part frm the research. The fllwing (table 2.) table shws the results, that prves the cnflicting assessments. The present part had nt any gals, because we can nt be sure which questins were answered (i.e. did the students rank hw the mentined situatins are likely r risky). Statistics Ethical Financial Health Scial Recreatinal N Valid Missing Mean 2,2069 3,1552 2,6207 3,7241 3,5862 Median 2,0000 3,0000 3,0000 4,0000 4,0000 Mde 1,00 4,00 3,00 2,00a 5,00 Std. Deviatin 1, , , , ,89957 Minimum,00,00,00 1,00,00 Maximum 6,00 6,00 6,00 6,00 6,00 a. Multiple mdes exist. The smallest value is shwn Table 2 Given answers fr the prblematical part f survey 2.2 Materials and prcedure The whle research was made at the Obuda University. Firstly we discussed the prblem and designed the survey with MsC students. After that, paper 163

4 questinnaires were given fr students during the first Ecnmic psychlgy lessn. This curse is in English fr Erasmus and Hungarian students. The instructins were general, and the papers were given persnally t the participants. All f the respnses were upladed in a table. The tw Hungarian and ne German answers were translated int English, and all f them were checked by spelling. Fr the evaluatin prcedures we have used (free) NViv cntent analysis sftware and SPSS. 2.3 Sample Because ur main cnceptin was bearing an extraplatin, s we did nt mnitr representative sample. Mrever this methd des nt exact representativeness. The frequency tables f the sample are fllwing: Gender Frequency Percent Valid Percent Cumulative Percent Valid Male 15 51,7 51,7 51,7 Female 14 48,3 48,3 100,0 Ttal ,0 100,0 Age Frequency Percent Valid Percent Cumulative Percent Valid 19,00 2 6,9 6,9 6,9 20, ,2 17,2 24,1 21, ,7 20,7 44,8 22,00 2 6,9 6,9 51,7 23, ,5 34,5 86,2 24,00 1 3,4 3,4 89,7 25,00 2 6,9 6,9 96,6 27,00 1 3,4 3,4 100,0 Ttal ,0 100,0 Natinality Frequency Percent Valid Percent Cumulative Percent Valid Armenian 1 3,4 3,4 3,4 Rmanian 1 3,4 3,4 6,9 Turkish 2 6,9 6,9 13,8 German 12 41,4 41,4 55,2 Hungarian 8 27,6 27,6 82,8 Dutch 1 3,4 3,4 86,2 Ukrainian 1 3,4 3,4 89,7 Spanish 1 3,4 3,4 93,1 Iceland 1 3,4 3,4 96,6 French 1 3,4 3,4 100,0 Ttal ,0 100,0 Table 3 Distributin f the sample 164

5 As we can realize the German and Hungarian students were ver represented and mst f the participants were 23 years ld. 3 Results Firstly we mnitred the meaning f risk, what kind f expressins came firstly int subjects mind. It is necessary t underline, that nn f the students speak English as native but they culd use dictinaries during the research. The cntent analysis based n tw stages: 1. autmatic cding regarding frequencies f every wrds 2. pen cding, which means that we read thrugh the text several times and started t create relevant grups. These cdes helped us t find sme cnnectins between the subsamples. As we have mentined earlier, the next part f the survey was nt able t be evaluated. Last part we used descriptive statistics because all f the respnses were measured n nminal scales. 3.1 Cntent analysis Autmatic cding Sftware f NViv cde the text autmatically which means withut any meaning nly the frequencies f the expressins will be cunted. Wrd Length Cunt Weighted Percentage (%) Similar Wrds risk ,16 risk smething ,08 smething decisin 8 9 3,67 decisin, decisins knw 4 7 2,86 knw, knwing danger 6 6 2,45 danger, dangerus situatin 9 6 2,45 situatin take 4 6 2,45 take, takes smetimes 9 5 2,04 smetimes negative 8 4 1,63 negative utcme 7 4 1,63 utcme always 6 3 1,22 always bad 3 3 1,22 bad 165

6 chance 6 3 1,22 chance depends 7 3 1,22 depending, depends fear 4 3 1,22 fear gd 4 3 1,22 gd life 4 3 1,22 life lsing 6 3 1,22 lse, lsing make 4 3 1,22 make, makes, making new 3 3 1,22 new things 6 3 1,22 things Table 4 Frequencies f used expressins The autmatic cding can be illustrated with wrd clud diagrams, the size f the wrd shws its regularity. Gap-filling wrds like smething are nt imprtant s we skipped them ut. Figure 1 Clud diagram f autmatic cding In sum this analysis indicates that the subjects judged risk as danger (bad r negative) decisin situatin r utcme. Often mentined expressins prvide a gd base fr pen cding. Cntent analysis slutins give us pprtunities fr deeper text mining withut any explanatins. The fllwing pictures figure ut the envirnment f the ften mentined terms. 166

7 Figure 2 Text searches regarding ften mentined terms 167

8 3.1.2 Open cding After several times f extensive readings we have fund the afrementined cde structure: targets f risk culd be divided int situatin (n=4) when the subjects defined risk as a situatin pssibility (n=5) when they described risk as a pssible chices f their life uncertainty (n=11) when they see risk as an uncertain prblem (like in study f Kahneman and Tversky) behavir (n=3) when they identified risk with a behavir last but nt least when the students gave nly an example (n=4). the directin f the utcme were defined in three different ways psitive (n=2) negative (n=11) r bth (n=14) the whle (altgether) meanings (essences f the definitin) were the next three sme student (n=6) cncentrated n the utcme f risk thers (n=7) fcused n the danger and the rest (n=8) highlighted the whle prcess f decisin. There was nly ne significant (p=0,95) cnnectins between cdes. In case f danger and negative utcme directins symmetric measures shwed psitive relatinship (Cramer s V= 0,558 p=0,029). In case cmparisn f the subsamples regarding gender nly ne majr difference culd be fund (see Figure 3). Wmen mentined uncertainty ftener than men. 168

9 Figure 3 Text searches regarding ften mentined terms 3.2 Frequency tables f aspects In the last part we have used frequencies and crss tabs analysis t determine surces and mtivatins f risk. Originally Weber and Blais (Blais & Weber, 2006) used multilevel mdeling, they investigated the risk the risk return relatinship between risk taking and risk perceptin. Their empirical investigatins prvided a multiple risk cnstruct which cntains three bservatin f risk. They fund 1. differences in the perceptins f the riskiness f risky chice ptins (perceptin) 2. differences in the perceptins f perceived benefits f risk (benefit) 3. differences in willingness t take part in a risky situatin (risk-taking). 169

10 DOSPERT Scale allws us t assess cnventinal risk attitudes (reprted level f risk-taking), perceived risk-attitudes (reprted willingness t engage in a risky activity) and utcme f risk (reprted value f taking part in a risky situatin). In ur study we were interested in which aspects will mstly influence students decisin in a risky situatin. The students needed t chse which ptins will impact n their decisin. The last table prvides us an verview abut the chices. Benefit Frequency Percent Valid Percent Cumulative Percent Valid yes 24 82,8 82,8 82,8 n 1 3,4 3,4 86,2 n pinin 4 13,8 13,8 100,0 Ttal ,0 100,0 Perceptins Frequency Percent Valid Percent Cumulative Percent Valid yes 13 44,8 44,8 44,8 n 8 27,6 27,6 72,4 n pinin 8 27,6 27,6 100,0 Ttal ,0 100,0 Risk-taking Frequency Percent Valid Percent Cumulative Percent Valid yes 19 65,5 65,5 65,5 n 4 13,8 13,8 79,3 n pinin 6 20,7 20,7 100,0 Ttal ,0 100,0 Table 5 Frequency tables f each aspects Maybe the rder f the questins influenced the assessments, but the ftenest influence-factr was the benefit (utcme) f a decisin. Cmparing the frequencies f generated pen cdes with DOSPERT aspects we have fund middle strng relatinships between danger and risk-taking (Cramer s V= 0,476 p=0,038). Accrding t gender there were n differences. Cnclusins Gal f this study is t have an verview (r a wrk definitin) what the students think abut risk. Cntent analysis caused an useful essence abut the risk s assciatins f ur sample. Aim f risk regarding ur sample is decisin in a pssible situatatin which can end in gd r bad results s it is dangerus. Directin: The current results suggest that risk means danger decisin fr the students. Only tw students evaluated risk as a psitive term. They identified risk is a chance r an pprtunity. Mst f the students mentined the bth sides f the pssible utcme that means risk is an ptin which has t be weight. Sme f the 170

11 students gave a cncrete example t define risk. DOSPERT scale cntains 30 example frm different life situatatin three times. Hwever, this scale is a validated slutin t measure risk. Our preliminary studies shwed that tgether 90 statements are difficult t handle. That is why we tried t evaluate the dimensins f the befre mentined scale. Fr that we asked ur respndents ranking the five different life situatins frm three different pint f views (see Table 1). Unfrtunately we culd nt be able t take in cnsideratin all f different dimensins f risky situatins. But we have fund that the final benefit f a risky decisin will influence respndents mstly (like utcme as a regularly mentined expressin in the definitins). Finally it is necessary nt skip ut persnality. We tried t take cnsideratin all limitatins f this research, in sum it culd be a gd base fr the future. One f the advantages that we had a feedback frm ur respndents definitins which shw sme similarities with the DOSPERT Scale. Acknwledgement We are thankful fr the suggestins f 2016/2017 autumn semester Research Methdlgy Students, and fr the answers f 2016/2017 Autumn semester Students f Ecnmic Psychlgy. This paper SUPPORTED THROUGH THE NEW NATIONAL EXCELLENCE PROGRAM OF THE MINISTRY OF HUMAN CAPACITIES Appendix Risky survey Sme backgrund questins: Yur Gender: Male Female Yur Age: I am. years ld. Yur natinality: What is risk? Hw can yu describe it? (Yu can answer with yur very first ideas, wrds which cme in t yur mind r yu can draw as well.) What d yu think which situatin is mre likely t happen t yu? Please rate separately all f them (extremely risky: 1. nt at all: 6, N pinin: 0) Ethical situatins like Leaving yur yung children alne at hme while running an errand. Financial situatins like Investing 10% f yur annual incme in a new business venture. 171

12 Health r Safety situatins like Riding a mtrcycle withut a helmet. Scial situatins like Chsing a career that yu truly enjy ver a mre secure ne. Recreatinal situatins like Bungee jumping ff a tall bridge. If yu need t value a situatin (regarding risk) which aspect influence yur decisin? (Y: yes, N: n, NO: N pinin) Expected Benefits f the situatins the benefits yu wuld btain frm each situatin. Perceptins f these situatins In this case each situatins have t be indicated (is the pssibility f negative cnsequences) hw risky yu perceive it. Risk-Taking : the likelihd that yu wuld engage in the described activity r behavir if yu were t find yurself in that situatin. Thank Yu fr yur answers! References [1] Blais, A.-R. & Weber, E. U., A Dmain-Specific Risk-Taking (DOSPERT) scale fr adult ppulatins. Judgement and Decisin Making, Vl 1. N 1., pp [2] Center fr Decisin Sciences, DOSPERT ORG. [Online] Available at: [Octber 2016]. [3] Faragó, K., Siker, kckázatvállalás és versengés a szervezetekben. Alkalmaztt Pszichlógia, pp [4] Kahneman, D. & Tversky, A., Prspect Thery: An Analysis f Decisin under Risk. Ecnmetrica, Vl 47. N 2., pp [5] Klnhfer-Derecskei Anita Reicher Regina Zsuzsánna (2016): GenYus - Y generáció az Y generáció szemével, Vállalkzásfejlesztés a XXI. században VI. pp [6] Tversky, A. & Kahneman, D., Judgement Under Uncertainty: Heuristics and Biases. Science, pp [7] Weber, E. U., Blais, A.-R. & Betz, N., A dmain-specific risk attitude scale: Measuring risk perceptins and risk behavirs.. Jurnal f Behaviral Decisin Making, 15., pp