Superreal Information (SRI)

A Grounded Theory Study on Content-Based Stimulation of Word-of-Mouth

Index

1. Preamble

The study discussed in this essay was part of a master's thesis in 'MSc Professional Practice [Creative Media Industries]' from the SAE Institute, in partnership of the Middlesex University London. The experiment (cf. 2.2) was executed in cooperation with the Department of Psychology of the SRH Heidelberg. The supervisor for the degree program was Dr. Fares Kayali. The assessors for the master's thesis were Dr. Ralf Muhlberger and Dr. Oliver Hödl.

Although this essay is a brief summary of the study and its results (outlined for open review), it is nevertheless detailed enough to comprehend the process of the study and its evaluation. To understand the key message of this study, it should be sufficient to read the preamble (cf. 1.1 ff.).
The original thesis was written in German and is available by personal inquiry.

1.1 Abstract

"Open your ears; for which of you will stop
The vent of hearing, when loud Rumour speaks?"

Shakespeare | King Henry IV, 2nd part, prolog

The outcome of the discussed study is an applicable model for content-based stimulation of word-of-mouth. In opposition to many recommendations which are published in this field, this model, as well as the theory on which it is based, does not merely suggest to "create interesting content for the right audience" or to "find the right communication channels e.g. in social media etc.". In fact, it should be assumed that a producer of creative contents will know how to create interesting content, for which audience it is appropriate and through which channels it shoud be communicated.
And most producers will know that it is not sufficient to merely have an interesting - or in any way "good" - content. Production and marketing are different kettles of fish. And above all, for unestablished producers it seems like they would have to create a nearly impossiblly high capacity of innovation, while established players seem to easily gain attention with even small features (Lutter, 2012).
But for all market participants, it is a crucial question as to how to evoke word-of-mouth, since in the end - even after all the advertising and public relation - word-of-mouth is the indicator for the position of a content for an audience.

While, for instance, the very recommendable paper of Mark Lutter (Lutter, 2012) highlights the social aspects and criteria for success within the creative media industries, the study discussed in this essay puts less focus on the successful person and more on the successful content. Nonetheless, both studies agree to a specific point, a specific premise, which is essential for understanding both the social/personal and the content-based aspects, and that is: Success is not related to quality. This premise is imperative for any theory which tries to find a reasonable commonality among successful contents and persons. Of course, this does not mean that there might not be a correlation between quality and success in regard to how the success is seen.

Having said that, the actual avail of the discussed study should be outlined: The Superreal Information Theory provides an extrapolation of the critical commonality of widely (trans-local) disseminated information, while the Indeterministic Information Theory describes the fundamental mechanism, with which the superreal dissemination can be explained. The derived model provides a method to apply the gathered perceptions in the professional practice of the creative media industries.

This abstract shall close with the obligatory attempt to sum up the exploration in one sentence (which admittedly is a bit cryptic in this case, but bridges neatly to the quote in the beginning): Success is just a rumour.

1.2 Object of research

As addressed in the abstract (cf. 1.1), the crucial problem the creative media industries have, is the fact that there are many recipients who ask for creative contents, but there are also many providers who satisfy that demand. In such a market which has an unmanageable amount of contents, one needs a high advertising weight to attract attention. And this translates into high costs and mostly high coverage waste. Word-of-Mouth (WoM) is supposed to be more efficient than confronting advertisement (Bernays, 2014). However WoM requires stable networks to run in. Nevertheless for creative professionals, WoM seems to be the right choice to communicate their contents to potential recipients at low costs. Logistically however, it appears that this form of information dissemination is very complex and one cannot communicate any content to any recipient.

The present study started with the question of how to disseminate information, and it led to the seemingly more important question of which information can be disseminated, and finally ended with the formulation of a use case for the solution to the question of how can information be designed, which can be disseminated.

1.3 Research objective

The primary objective of this study was the investigation of possibilities to instrumentalize word-of-mouth. The secondary objective was to situate the resulting application models on theories, which are congenial to further investigation.

1.4 State of research

Word-of-mouth is researched largely within the disciplines of communication science and marketing research. While the latter dicipline suggests more decided models and explains the phenomenon in a more differentiated manner, the former has more of a general view on the phenomenon and sees it as a sub-category of common communication (Lis and Korchmar, 2013). But due to the strong focus on product sales, the marketing research theories often show a lack of theoretical explanation power.
Developed with the Grounded Theory Method, the present study attempts to supply a completely new approach and offers a theory which can be used as a base for further investigation in both disciplines.

(In order to keep this document short, it should be noted that further information about the state of research in regards to this topic can be found in the reference list (e.g. Oetting, (2009).)

1.5 Method of research

The method of research was compliant to the requirements of the Grounded Theory, which has been developed by Glaser and Strauss in the 1960s (Glaser and Strauss, 1998). The aim of the method is to develop theories from research concepts. According to this, the theory is not the starting point of the research, rather the theory arrives at the end, which means the theory is the result of the research.

In the course of the study, at first there was an initial concept: a technical concept for information dissemination (the so-called potage-project). This (mostly web-based) concept should allow to pass-through information through a user-based filter to recipients, who could be interested in the information. Based on this, the participants were interviewed in regard to their user-behavior. In doing so, the study was concerned with which were the reasons and motivations that made the recipients pass-through a specific information compared to another. As a result, a synthesis could be constituted, which disclosed the fundamental difficulty of such an information filter. Following this, a commonality of the passed information was extrapolated. This resulted in altering of the subject of investigation, which then subsequently became the question about the content, instead of the (technical) method. Therefor, a hypothesis was built which describes the trans-local dissemination of superreal information. Based on this, an experiment could be configured. Due to this, a (content-based) variable was derived from the hypothesis, which should be able to influence the probability of the dissemination of an information. The following experimental set-up tested the variable on probands. In the process, variegating contents were communicated towards the probands and the dissemination of the information was observed. The evaluation afterwards allowed to postulate a theory about indeterministic information. Finally, an applicable model was derived from the theory.

2. Trails

There had been two major trails within the study: Firstly the testing of the potage and interviews with the participants, and secondly the experiment in regard to the postulated variable.

As observed across most existing literature on this topic, every epidemic information dissemination process in the end is a word-of-mouth effect (Röthlinghöfer, 2008), and thus word-of-mouth was chosen as an indicator for the experiment. The main problem was to excogitate an experimental set-up, in which the act of observing did not contaminate the actual moment of dissemination. It was considered that perhaps this can be achieve by using the technical possibilities of the potage-project. However due to the uncommon nature of a potage, the conceptual confusion would have made the trail inaccurate; so the potage-project was inapt and an in persona trail was the method of choice.

2.1 Initial concept - potage

The potage-project tried to offer a technically supported way to pass-through information only to those participants who were interested in it. Thereby, the group of disseminators should act as the filter itself. This should be achieved on the one hand, with the actual information to disseminate (the content) was "labeled" with a description of itself. (This description or label may not contain any advertising formulations.) On the other hand, a receiving participant was obligated to pass-through such a labeled content (= a potage), particularly when he or she is personally NOT interested in the content, judged by reading the description - in the best case to somebody, who may be interested in it.

[Ill. 1.: potage concept example]

To keep this document short, there will be no explanation about the potage-project in more detail. Further information about it as well as the beta-version (May 2016) is available here:

http://potage.openpot.com/

An exemplary potage is available here:

http://openpot.org/29uxEMRT

2.1.1 Results

With version 0.0.6-alpha of the potage-project, the first field tests were started. Therefore, a set of potages with specialized topics for the most part, was designed. While disseminating this potages it was important to instruct a section of the participants to pass-through the potage unopened in any case (instruct them to be a generation zero seeder), in order to randomize the first generation more effectively.

The tracking could be achieved by use of the potage's immanent tracking function and the server log-files. The potage database supplied the chronological distribution of the accessing participants.

For the evaluation, the most important question was that, how many generations did a potage survive (Sn), followed by the variable whether in the next generation more or less or equal disseminators existed (En%). The variables of how many disseminators opened a potage (On%) or pass-through (Fn%), was allocated for the creation of the initial condition (premise). Results for a successful running project would have had values of Sn >= 5 and En > 100%.

Result:
Sn <= 1,2
En <= 10%

The average result of the potage tests was, that even the second generation could not be built. Practically all potages - regardless how large the first generation had been - were not able to generate further disseminators in the second generation.

In order to infer the initial condition, it was necessary to derive from On% and Fn%, as to how the behavior of specific participants affected the process:

Result:
On >= 98%
Fn <= 2%

In conclusion it could be asserted, that practically all disseminators of the first generation opened the potage they received regardless of how specific (and therefore only for a specific target group) the received topic had been designed. Thus, the potage could not be disseminated in a new generation.

Therefore the potage-project does not work as expected, since every disseminator opened the potage and because of this a new generation of disseminators couldn't be built.

2.1.2 Interviews

To find an explanation for the user-behavior of the potage participants, a qualitative survey was conducted with random users (ca. 15%). Whereever possible a face-to-face conversation was undertaken. A small number of participants were interviewed via e-mail (ca. 5%). The questions were not standardized, but every interview surveyed at least the following aspects:

- Was the potage-concept understood?
- Was the usability easy?
- What was the reason for a decision to open or pass a potage?
- What is the personal opinion about the potage-project?
- How would a further user-behavior look like?

The analysis of the answers focused primarily on the question of why almost every participant always opened every potage. Thereby, an ex ante thesis was tested:

Thesis:
Because the potage-concept is new, the participants are not used to pass-through contents, which they did not examine earlier. Also, they are not really aware that once they opened it, they cannot pass-through it any longer.

Based on the collected data, this thesis had to be dismissed. The survey could not show that the participants weren't aware of losing the possibility to pass a potage when opened, nor could it be shown that the novelty of the project would correlate with the decision to check the content in any case. The analysis of the data thereby, led to another assumption:

Antithesis:
The interest, which results from the fact that a potage was sent by a friend or even a stranger, is more important for the decision to open a potage, than the knowledge / wish to keep the potage closed, due to general disinterest.

Most of the participants could specify that they opened the potage because they wanted to know at all costs as to why the sender of the potage could have had the opinion that this potage could be interesting specifically for them. As this antithesis does not completely exclude the thesis, it was possible to build a synthesis:

Synthesis:
The interest in a content is influenced more by the sender than by a neutral and impersonalized description of the content itself. However, this could change theoretically when users get used to the potage-concept and the personal interest lowers, due to frequent use, and therefore, the description gets regarded more.

In fact this means, that the potage-concept does not have any realistic chance to become a popular way of information dissemination. Because in order to surmount the initial barrier, it would have to be already popular, which it cannot become, because the initial barrier cannot be surmounted.

2.2 Experiment

The purpose of the experiment was the empirical examination of a theoretically prognosticated variable for content-based stimulation of word-of-mouth.

2.2.1 The variable

The theoretical prognosticated variable PA describes the relation of two sub-variables:

P = Plausibility
A = Absurdity

Plausible means the kind of information, which a rational thinking individual will successfully match with learned and experienced knowledge. Absurd means information, which will contradict the knowledge of such an individual. For instance plausible means, that a stone which is thrown in the air will eventually fall back on earth. Absurd is, that it transforms into a chipmunk, which will reach parabolic speed and build a nest on moon. But plausible does not have to be necessarily true or according to facts. Neither does absurd have to be necessarily untrue or contradictory to facts. For instance, most people would consider it as plausible, that a dime which is thrown from the Empire State Building, will speed-up while falling to a point where it can cause lethal injuries if hitting somebody on ground. But in fact New York would have to be in a vacuum, that the coin could accelerate so much speed.
This means, the two terms relate to the personal estimation, if an information is true or untrue. Therefore, it varies from person to person.

The variable P/A or PA describes the relation of P to A. Since a single information can hardly include both interpretations simultaneously, it is clear that the PA-variable always relates to a context or conglomeration of information, thus a story. A story can include plausible and absurd aspects. A story can result into an overall interpretation - a conclusion - which is more complex than a single information, and therefore can include both aspects simultaneously. For instance, the following could be a story:

A man walked into a store and wanted to buy a watch. But the saleswoman did not sell him one, because he had no money with him.

This story appears to be completely plausible. Most - maybe even all - readers have learned, that one can only purchase a product, if one can surrender a counter-value, for instance in form of money. The PA relation is therefore strongly on the plausible side (PA = 0). However the story could also be completely absurd:

A man walked into a store and wanted to buy a watch. Because he had no money with him, the saleswoman gave him a pair of yellow shoes.

This story appears to be absurd, because of the causal connection between the fact that the man had no money with him and the giving of the yellow shoes, does not make sense. The PA relation is therefore strongly on the absurd side (PA = 1).
(Atleast for most people. Let's consider a society, where it is perfectly normal to receive yellow shoes when a person is illiquid. In this case, a member of that society would consider the story as perfect plausible.)

A story can also have a more equally weighted relation of PA. For instance:

A man walked into a store and wanted to buy a watch. However the saleswoman did not sell him one, because he was wearing yellow shoes.

On the one hand this story seems to be absurd, since one has hardly heard about somebody being unable to buy a watch because he was wearing yellow shoes. On the other hand, it seems to be plausible, since at least one has heard that one does not get entrance to specific locations, when wearing the wrong footwear. So it would be imaginable, that this also could be applicable to a store with (maybe) expensive watches. The PA relation is more or less equal (PA = 0.5).

The PA variable is not quantifiable. This is due to the subjective interpretation of P and A and also because there is no quantifiable scale for how strong an information contradicts learned knowledge, or not. But in order to have a usecase for the variable, it is sufficient to approximately define a relation between 0 and 1 (where 0.5 represents the "ideal" equation). Further in this essay, for the sake of convenience, "stories with the variable" and "stories without the variable" will be discussed, where with implies a PA value around 0.5 and without either a value around 0 or 1.

2.2.2 Thesis

Thesis:
A story with an equally weighted PA factor has a significantly higher probability to get retold than a story with an extreme PA factor.

Contradictory to the popular opinion within extant literature, that absurd stories are superior to plausible in regard to their memorability and therefore retelling, the thesis to this experiment claims that essentially the probability of retelling increases towards the equation of both factors, and decreases at the extremes.

[Ill. 2.: assumed distribution PA]

2.2.3 Experimental set-up

The experiment was executed in three steps. The first step was the telling of test stories to the probands and the simultaneous psychological observation of their behavior. The second step was giving the probands the opportunity to retell the stories, without them knowing that this would be part of the experiment. The third step was the measurement of the retold stories.

2.2.3.1 First step:

[Ill. 3.: experimental set-up 1. step]

In the first step two probands in two groups were "interviewed" by respectively one administrator at a time and observed by respectively one observer. The probands were instructed by the administrator, that ten short stories will be narrated to them. After each story they were obligated to give a short comment in regard to the story. They could speak freely and without saying anything that could be presumed as being wrong, and in case they wouldn't have any opinion, that would also be a valid answer. The administrator was instructed not to influence the dynamics between the two probands and not to react to the comments as far as possible. After the comments were given, he just shall proceed to narrating the next story.

The observer was facing the administrator, in order to refract the attention from him. The oberservers' task was to observe the probands' behavior after they heard a story and to record the observations. Thus, the main focus was on the observation of reconciliatory behavior as a result of uncertainty. As reconciliation was anything considered, which represented a casual exchange of information. However, this had to be based on uncertainty, respectively a previous uncertainty had to occur.
There were several indicators defined, in regard to what as an uncertainty had to be observed: (This list is shortened and detailed explanations are missing in this essay. Please refer to the reference list for further details.)

- To wind the feet around the chair legs (Navarro, 2008)
- Disappearing thumbs (Navarro, 2013)
- To chew one's fingernails (Navarro, 2008)
- Pacification gestures / deflectors (Ekman, 2007)
- Forward leaning body (Ekman, 2007)
- Low / murmurous voice (Navarro, 2008)
- Dipping shoulders (Ekman, 2007)
- Scratching the neck / side of the head (Krüll, 2012)
- Biting the bottom lip (Darwin, 1872)
- Etc ...

In general, it was important to pay attention that other causes for these behaviors could be excluded. For instance, could dorsal pain be the cause for a forward leaning posture, or frequent winking could be caused by dry ambient air, etc.. Therefore, the relative peculiarity was crucial. It was recorded on which stories the probands behaved differently, in relation to the prepended indicators and their common body language.

2.2.3.2 Second step:

[Ill. 4.: experimental set-up 2. step]

In the second step the actual generation of the relevant data occurred. Therefor, two parallel interviewed groups at a time were brought together in a separate room immediately after the interview. They were instructed to wait there on the pretext that a short pre-evaluation would be necessary. After this, there would be a last question which would be asked to all four of them. The probands were left in this separate room for a while, and had the possibility to see the administrators through a glass door doing their pretended pre-evaluation. The administrators (or also observers) were instructed to pretend to compare their data in a calm way and without attracting attention. The purpose of this was to put the probands in an isolated situation in which they felt unobserved (which they were in fact). But they should also be able to see that the experiment "is still running". This should open a frame in which the probands could talk about the experiment, without having the feel to still be part of it. For this reason, the probability should be quite high, that both groups exchange their experiences with the experiment among themselves. Consequently, this provided the conditions for the crucial question which stories they would retell, without being influenced - not even by an observation - or being instructed to do so.

2.2.3.3 Third step:

After the time had elapsed, the four probands were asked the last question. This last question was the actual core of the experiment. The question was, which stories of each group the other group knew in the meanwhile. Every retold story got recorded. Therefore, it was irrelevant if a story was retold correctly or completely. If the repetition of the story's content was sufficient to recognize the story, it got recorded. The reason is the perception, that there is no demand for exactitude in a retelling process (Möller, 2011). Word-of-mouth is not a medium as defined by media science. Rather it is a dynamic process which does not attach importance to exact reproduction of an information's form.

2.2.4 Parameters

2.2.4.1 Randomization:

Every kind of proband, irrespecitve of gender, age or origin (social milieu) was applicable. However, for logistical reasons, all probands were recruited in proximate periphery of the campus of the SRH Heidelberg. According to this, the largest fraction was young adult students of both(/all) genders. Because of this, the probands are not representative for a whole population. In spite of the local connection, the probands still represent valid samples, because the human handling of information is either so universal or so individual, that a limitation to participants out of a specific milieu should not affect the results (Watzlawick, 1988).
The groups (duads) were built as heterogeneous as possible, so that preferably friends / strangers, of same sex / different sex, of same age / different age etc. were evenly distributed.
The order of the told stories, which were narrated by the administrators, was randomized in advance by use of the Mersenne Twister Algorithm, so that the position of a story within the order would not affect the results.

2.2.4.2 Double-blind:

The experiment was designed to be double-blind. The probands did not know what was researched in particular. Primarily in step two, they acted on the assumption that the experiment would already be finished, and that their current behavior wouldn't be observed and therefore, would not have any impact on the study.
The administrators weren't informed about the nature of the experiment. Primarily, they did not know in which way the read out stories differed, and therefore could not signalize a specific expectation to the probands to react to specific stories in a different way than to others.
Also, the observers did not know how to differ the stories. Thus, they could not observe and record specific stories "more complaisant" than others.

2.2.4.3 Control group:

Due to the logic of the experimental set-up, a "real" control group would have been pointless. Because a group that only gets told stories without the variable, could afterwards only retell stories without variable. The purpose of the experiment was not to investigate if there would be retold stories at all, but which stories would be retold. Therefore the priciple of a control group was achieved by the idea, that the one group (in a setting of two parallel interviewed groups) can function as a control group for the other and vice versa. This could be assured by telling different stories with variable to both groups, while the stories without the variable could repeat. In case there would have been an alternative factor which influenced the motivation for a retelling among the stories with variable, the other group would not have had heard this story and couldn't retell it in any case. So, if the prognosticated variable would be wrong, there would appear a significant occurrence of specific stories with variable. Based on this unequal distribution a different variable than the prognosticated would have to be assumed.

2.2.4.4 Quantification:

A total of 240 stories were told to 48 probands in 12 groups in 2 sub-groups at a time. 3 out of 10 stories of 240 contained the variable. 168 stories were without variable and 72 with variable. The stories with variable rotated in a period of 12, stories without the variable within 2 sub-groups in a period of 7.
5 out of 12 (ca. 41%) of the probands were female. The average age was ca. 23 years, within a scope from 18 to 32 years.
The experiment was conducted on 3 working days within 1 week, between 16:00 - 20:00 houres in silent learning rooms of the SRH Heidelberg. A single cycle of the experiment took ca. 20 - 30 minutes, while the first step took ca. 10 minutes, the second step ca. 10 minutes and the third step ca. 5-10 minutes.

2.2.4.5 Story design:

The stories were partly self-written and partly modifications derived from external sources (Brednich, 1990 - 2004). It was assumed that these stories from external sources - for instance urban/modern legends - principally must fulfill the criteria for an equally weighted PA factor - otherwise they couldn't have disseminated (cf. 3.2). A part of the stories with an extreme PA factor was designed by removing the ambivalent aspect from functioning stories/legends. It was important to pay attention that all stories were quite similar to each other, in regard to their length, scene and protagonists, as well as in the word choice / diction and text design. The more analog the form of the stories was, the less the probability that a significant frequent dissemination of a specific story had different/alternative reasons than the prognosticated variable.
As an example, the following is one story with variable and one without variable:

A:
A couple from Canterbury was on vacation in Tokyo. On the last day they wanted to explore the city on their own. Because they weren't able to speak Japanese, they photographed street signs which appeared relevant to them, to be on the safe side in order to find their way back home. Later however, nobody was able to help them with their way home: they only photographed signs with Japanese characters which meant "one-way street". They weren't able to find their hotel in time and missed their flight.

B:
In Oxford a woman wanted to purchase a nice dress for herself. She searched for quite a long time and finally found a lovely evening dress in an expensive boutique. Although it was bit pricey and also required a little modification, she decided to buy it. She ordered the modification, which the boutique was willing to do, and a few days later she picked up the dress. At home she did not like the dress anymore and she returned it. The boutique rejected the return due to the specific modification for her.

The obvious similarity of story A and B shows, that the prognosticated variable is in the content - not in the form.
As a remark it should be mentioned, that the interpretation of the P and A sub-variables is a subjective process. It will not be interpreted - also not in regard to these two stories - in a same way by every reader. The experiment aimed to show, if it is possible to design a story in such a way, that a specific interpretation occurs in a significant frequency.

(A is IDI, B is not (cf. 4.2))

2.2.5 Evaluation

(In this essay only the evaluation of the data of the third step is shown, as this data is more relevant in regard to the derived theory and to keep this essay short. Further information is available in the original master's thesis.)

During the evaluation the statistical significance against the alpha error got calculated. Therefore the correspondent null hypothesis was formulated and the level of significance defined.

Null hypothesis:
The distribution of the test results occured at random.

The level of significance was set to:

alpha = 5%

(Information about the reasons for the levels of significance can be found in the original master's thesis.)

Out of 240 stories told in step 1, 39 were retold in step 2 (ca. 16%). Thereby 19 stories did not contain the variable (ca. 49%) and 20 did contain the variable (ca. 51%).

Story Av Bv Cv Dv Ev Fv Gv Hv Iv Jv Kv Lv A0 B0 C0 D0 E0 F0 G0
Retelling 2 2 2 2 2 1 2 2 2 1 0 2 2 5 4 1 4 2 1
Sum 20 19
[Tab. 1.: data third step]

Out of 240 stories told, 168 (70%) did not contain the variable and 72 (30%) did. 19 out of 168 stories equals ca. 11.31% and 20 out of 72 ca. 27.77%.

Stories which did contain the variable were retold 2.4553x more often than stories which did not contain the variable.

For the calculation of the null hypothesis for the alpha error this means:

Probabilities (binomial dissemination)

of      n = 39 samples
with    k = 3.3333333333333 options (10:3)
to hit  m = 20 times
with    r = 1 alternative/s
using   totheright method
at      alpha = 5% significance niveau

n = 00 -> binom. coeff. = 00000000001 -> P = 00.0001% - 
n = 01 -> binom. coeff. = 00000000039 -> P = 00.0015% - 
n = 02 -> binom. coeff. = 00000000741 -> P = 00.0124% - 
n = 03 -> binom. coeff. = 00000009139 -> P = 00.0654% - 
n = 04 -> binom. coeff. = 00000082251 -> P = 00.2524% - #
n = 05 -> binom. coeff. = 00000575757 -> P = 00.7571% - ###
n = 06 -> binom. coeff. = 00003262623 -> P = 01.8388% - #######
n = 07 -> binom. coeff. = 00015380937 -> P = 03.7151% - ###############
n = 08 -> binom. coeff. = 00061523748 -> P = 06.3687% - #########################
n = 09 -> binom. coeff. = 00211915132 -> P = 09.4014% - ######################################
n = 10 -> binom. coeff. = 00635745396 -> P = 12.0876% - ################################################
n = 11 -> binom. coeff. = 01676056044 -> P = 13.6574% - ####################################################### <- median probability
n = 12 -> binom. coeff. = 03910797436 -> P = 13.6574% - #######################################################
n = 13 -> binom. coeff. = 08122425444 -> P = 12.1566% - #################################################
n = 14 -> binom. coeff. = 15084504396 -> P = 09.6756% - #######################################
n = 15 -> binom. coeff. = 25140840660 -> P = 06.9112% - ############################
n = 16 -> binom. coeff. = 37711260990 -> P = 04.4429% - ################## <- alpha
n = 17 -> binom. coeff. = 51021117810 -> P = 02.5761% - ##########
n = 18 -> binom. coeff. = 62359143990 -> P = 01.3494% - #####
n = 19 -> binom. coeff. = 68923264410 -> P = 00.6392% - ###
n = 20 -> binom. coeff. = 68923264410 -> P = 00.2739% o # <- test result
n = 21 -> binom. coeff. = 62359143990 -> P = 00.1062% + 
n = 22 -> binom. coeff. = 51021117810 -> P = 00.0372% + 
n = 23 -> binom. coeff. = 37711260990 -> P = 00.0118% + 
n = 24 -> binom. coeff. = 25140840660 -> P = 00.0034% + 
n = 25 -> binom. coeff. = 15084504396 -> P = 00.0009% + 
n = 26 -> binom. coeff. = 08122425444 -> P = 00.0002% + 
n = 27 -> binom. coeff. = 03910797436 -> P = 00.0000% + 
n = 28 -> binom. coeff. = 01676056044 -> P = 00.0000% + 
n = 29 -> binom. coeff. = 00635745396 -> P = 00.0000% + 
n = 30 -> binom. coeff. = 00211915132 -> P = 00.0000% + 
n = 31 -> binom. coeff. = 00061523748 -> P = 00.0000% + 
n = 32 -> binom. coeff. = 00015380937 -> P = 00.0000% + 
n = 33 -> binom. coeff. = 00003262623 -> P = 00.0000% + 
n = 34 -> binom. coeff. = 00000575757 -> P = 00.0000% + 
n = 35 -> binom. coeff. = 00000082251 -> P = 00.0000% + 
n = 36 -> binom. coeff. = 00000009139 -> P = 00.0000% + 
n = 37 -> binom. coeff. = 00000000741 -> P = 00.0000% + 
n = 38 -> binom. coeff. = 00000000039 -> P = 00.0000% + 
n = 39 -> binom. coeff. = 00000000001 -> P = 00.0000% + 

p-value for n = 20 (to the right) = 0.4336%
[Tab. 2.: significance calculation (tab. 1)]

The probability, that the test result occurred at random, in case the null hypothesis is true, is ca. 0.4%. Since the level of significance was set to 5%, the null hypothesis can be dismissed.
In spite of the result that the value of significance undercut the level of significance, merely a correlation between the prognosticated variable and the retelling can be presumed, but not imperatively a causality.

(The software which calculated the results and supplied the above output (cf. tab. 2) was especially made for this study and published as free and open-source software:
http://solutions.openpot.com/binomnom/index.php )

For the interpretation of the results, one should not forget that the significance relates to a statistical average. Therefore it cannot be observed with any story and with any proband under all circumstances. Nevertheless it can be eliminated that the distribution of the told stories occurred at random for the benefit of the stories with variable.

Ergo:
There is a coherence between the design of the PA variable and the retelling probability.

(Please refer to the original thesis for information about the observed probands' behavior and its correlation to the retelling probability.)

All observations of stories with variable have in common that there seem to be a requirement for the probands to maintain a casual and interpersonal alignment. The behavioral and psychological evaluation of the observations (1. step) produced the conception that the variable promotes a uncertainty, which leads to an increased necessity for casual alignment with other people.
But for what purpose? Here a comprehensive term could be introduced: determination. In order to be able to give a - subjectively deemed as qualified - comment (cf. 2.2.3.1), first of all an information has to be determined. In this process, the attempt is to sort out the absurd (respectively superreal (cf. 3.2)) aspects - in case they exist. An information on which this trail leads into a circular dead-end can be called: Indeterministic Information (IDI).

3. Superreal Information

The requirements for the Superreal Information Theory were initially to be able to explain why a specific information gets disseminated, and not another. Furthermore, the theory should be applicable and most notably falsifiable.

3.1 Grounded Theory origin

During the evaluation of the potage data, an unexpected aspect stood out: A small part of the potages generated a 2. generation, in spite of the fact that all participants of the 1. generation did open the potage (cf. 2.1).
In order to review the possibility that this may be ascribed to an imprecision of the investigation of the generations, the spread potages had been examined in a cross-check. In doing so, it was noticed that this "tunnel effect" seemed to repeat itself among specific potages. An interrogation of involved participants was able to confirm that specific potages were forwarded despite the opening (and despite the additional effort, since a potage is not meant to be forwarded when opened) but was not able to answer why and why these potages in particular.
A hermetical inspection of the tunneling potages led to a radical consideration and the change of the intended system. Since the Grounded Theory Method allows this change, or even demands it, it was undertaken to extrapolate a commonality of the few data sets of the tunneling potages, by use of the comparative method. So, the attention was turned to the question of which information gets disseminated, instead of who disseminates how.

Due to the fact that at this point of the Grounded Theory there was no initial concept but a question derived from the intended system, it was obligatory to develop a thesis or model or theory for the further research.

3.2 SRI Theory

Since the problem was related to a new intended system, it was not required anymore to restrain only to potages. Popular and unpopular examples of successful (widely spread / known) and unsuccessful information / products / persons / events were added, from the fields of art, culture and science. (For instance: Orson Welles/H.G. Welles „War of the Worlds“, J.W. Goethe „Werther“, Markus Persson „Minecraft“, Evan Goldberg „The Interview“, Andrea Rossi „E-Cat“, CERN „CNGS-Neutrinos - Gran Sasso“, Biographie „Vincent van Gogh“, Biographie „Priscilla Ann Wagner“, Uri Geller „Drei mal Neun (1974)“, Daniel Myrick „Blair Witch Project“, Jan Böhmermann „#varoufake“, Reyhan Şahin „Lady Bitch Ray“, Mythos „Coca Cola zersetzt Fleisch“, Meme „Informed Techniker“, Apple „MacBook-Air Marketing“, ZDF „Schokolade macht schlank“.) Some were better documented and some worse, in a historic or contemporary manner. Subsequently, these stories were examined in regard to the from the tunnelling potages extrapolated variable. Consequently, a model / theory was constituted as well as the predication as follows:

An unreal information is an information which contradicts the as real considered experience. The radius of dissemination of a real information surmounts hardly the local environment, whereas an unreal information can gain a trans-local dissemination. The effect of a trans-local dissemination of an unreal information is that the information will be accepted as real after the dissemination. The trans-local dissemination causes a transformation process. In this process an initially unreal information gains a real character. The information becomes superreal.

[Ill. 5.: superreal dissemination]

This is to be illustrated with a case example: On October 30th 1938 the American radio station CBS broadcast an audio drama made by Orson Welles, which was based on the book "War of Worlds" (H. G. Wells). The science-fiction novel was re-enacted, based on an adaption of Howard Koch. Furthermore, the location was replaced from England to the USA and the story was modified accordingly.
According to media reports afterwards, the audio drama provoked a mass panic among the citizens of New York and New Jersey, because many listeners were unable to realize the fictitious nature of the audio drama and acted on the assumption that America would be attacked de facto by extra-terrestrial invaders. According to other reports, this panic had spread even across the whole country. This story became famous worldwide, in spite of the fact that there has been absolutely no evidence for the claim of such a mass panic. The specialist literature in the field of communication science assumes that the report was a hoax (Jefferson and Socolow, 2013). Today one would call it: successful buzz-/guerrilla-marketing.
Nevertheless, the story is pervasive and well-established and is considered to be absolutely true by many people. One could say: it gained some entitlement for truthfulness. Even though a rational thinking individual could easily come to the conclusion that everyone should be clear on how unlikely an alien invasion is, and this should also be valid for the people from 1938 in America, since why should they specifically tend to believe something like this more than others.

On closer examination of other very famous stories about events or "facts", like for instance that Goethes' "Suffering of the Young Werther" caused epidemic suicide after it was published or the persistent myth that Coca Cola would be able to degrade meat and bones, one could conclude that a superreal moment is maybe inherent in every trans-locally famous story.
However the differentiation of the superreal variable shows some limitations:

- Superreal does not mean imaginary
- The superreal moment is not re-constructible with any famous story
- Vital information is excluded (e.g. the message about a natural disaster)

(The classification of superreal information with trans-local dissemination and real information with local dissemination will not be discussed in this essay. Please refer to the original thesis for further information.)

4. Indeterministic Information

When an information, which gained a wide dissemination and thereby a claim for trueness, can be called superreal, then the indeterministic information (IDI) is its precursor. An indeterministic information + its wide dissemination = a superreal information. The Indeterministic Information Theory shall provide a possible explanation for the implied phenomenon.

4.1 IDI Theory

The impulse to disseminate information is assessed biologically (Watzlawick, 1988). For a group of living individuals (e.g. in a natural environment), it is important to share information. Thereby, there has to be information which is important for only a few, and information which is important for most or all members. An information which would be important for all could for instance be, that the group is in danger. This information must be disseminated as fast and effectively as possible, in order to produce e.g. a mass-escape. The most natural form of information dissemination is to pass information from individual to individual, respectively from mouth to mouth. But it is initially irrelevant if this happens verbally or non-verbally. However, a group is not able to react to an optional amount of information. A permanent occupation with redundant information would paralyze the group in its capacity to act. There must be an intelligent filtering mechanism, which reduces the amount of disseminating information. For this purpose, information has to be determined. Each member of the group functions hereby as a filter and decides if a chain of dissemination for a specific information commences or not. This takes place by the simple decision to share or not to share an information. (This does not imply that this decision has to be made consciously.) If there is a critical mass of individuals who decide to share the information, the information becomes epidemic.
The crucial question now is which criteria exists for the decision-making. Therefor and first of all, there must be categories for information to pass and information to filter. In case an information matches with one of these categories, the information is determined and it can be reacted to it accordingly. This process can be diffuse. For instance, if an individual is not sure about how to determine an information, it can be passed to another one, in order to see if the other individual might be able to determine it. But what happens if nobody can determine the information? In its search of determination, the information gets disseminated until either it will get determined or all members of the group know it. An information which everyone knows but nobody could determine, determines itself due to its dissemination, and thereby gets its superreal character.

For what could this mechanism be useful? Let's assume for instance, that in a natural environment, a living group gets confronted with an unexpected alteration of their environment. For instance, suddenly a significant alteration of the geological, climatic or natural conditions occurs. An individual of the group, which comes across this alteration, will not be able to classify it and therefore will pass this information to another member of the group, because of the personal uncertainty of it. Since this other member also cannot classify the information, it will be passed to another again, and so on and so forth. Finally, once the whole group receives the information, the information has to get determined, because it cannot occupy the group endlessly. The alteration will ultimately be accepted as et est sicut est.

The designed indeterministic information "utilizes" this effect. The sudden and unexpected alteration represents the absurd aspect. The fact that this alteration does exist nonetheless represents the plausible aspect. If an information is able to cause a comparable uncertainty when being received, there is a high probability that the information will get disseminated. However a simple information, which e.g. produces fear, does not come into consideration, because this kind of information can mostly get determined, and there are specifications for the re/action. And these specifications will tend to not share the information in case of doubt. At least nowadays, because of the extremely high appearance of information, the benefit of filtering seems to be mostly higher than the benefit of sharing.

Indeterministic information is inconsistent but reproducible. The design of an information which a large ratio of recipients cannot classify, is always relative to the existing and learned structures of those recipients. Therefore an objective and always effective indeterministic information cannot exist. Though due to this, indeterministic information may repeat.

For the dissemination of an indeterministic information the disseminating group can have any size. Theoretically, this information could be so relevant, that all members - maybe not only of that group, maybe of the whole species - would be concerned. Here the indeterministic information differs from the "normal" deterministic information. An effective filter for normal information is the size of the group, within which it can disseminate. Such normal information loses relevance with each generation of dissemination (not: biological generation). Because of this, such information is mostly localized and personalized. Mr. XYZ did this or that. This information gets more and more uninteresting, the less a recipient still knows who Mr. XYZ even is. Therefore, one should strictly detach local and trans-local dissemination. For a local dissemination, indeterministic information is not necessary at all. A local food-store with good products in a small town can establish locally without any problems, without using any absurd aspects in its public communication. For the consideration in which cases indeterministic information is appropriate in the end, a deeper investigation with the identifiability of local and trans-local is obligatory - especially in regard to technical (mass-)media etc..

There are more (partly unclear) aspects, like e.g. the occurrence of dissemination dead-ends caused by diffusion of responsibility or the influence of personalized profiles on the motivation basis of word-of-mouth participants etc.. But these cannot all be discussed in this essay.

4.2 IDI Model

There shall be a model constructed for the utilization of the theory for the professional practice, that decisively enables the design of an indeterministic information and allows a supplementary controlling. The intended objective here is always the trans-local dissemination.
One of the most difficult aspects of the model is the fitting of the IDI with the advertised product. It is not possible to disseminate a random IDI, but it must relate in any way to the advertised product. The closer the story is to the product, the more likely the consumer will find the way to the product. But coincidentally one has to assume, that the closer the story has to be to the product, the less opportunities one will have for successfully designing the IDI. Therefore, one can expect that there are products with a higher potential for the fitting and products with a lower potential. Roughly estimated, one can say that creative products, e.g. artworks, include a higher potential than e.g. items of daily use. Nevertheless, there are examples of such products, to which an IDI successfully was attached (e.g. energy-drinks which have the reputation to contain extreme high doses of caffeine or traces of cocaine). For the fitting it is obligatory to carve out the product's appropriate plausible aspects. Every product will include some aspects, which are appropriate to a greater or lesser extent. This is a common marketing procedure, which can be recapped with the term 'USP' (unique selling proposition). Once a set of appropriate plausible aspects gets compiled, at least one of them must be expanded with an absurd component. For this purpose the absurd prototypes are helpful. Absurd prototypes are a set of bipolar basic patterns, on which most IDI is based. These narrative archetypes repeat constantly and get translated into the contemporary frame, solely by the plausible aspects. (The decoding of the prototypes would go way beyond the scope of this essay. But as an example: One absurd prototype can be for instance, a nearly impossible catenation of accidents leads to an implausibly great misfortune. The opposite pole would be the implausibly great fortune.)
The fitting has to be between the absurd prototype and the plausible aspect and the absurd prototype and the product. Not every IDI which got constructed of a plausible aspect and an absurd prototype, results in a good match, which one can confidently link with the product. So, it must be assured, that the IDI is in accordance with the product's aim and image.
In this process, the wished target group also has to be considered. In regard to the dissemination of IDI, it must be respected that the information will not necessarily spread only among the target group. Quite to the contrary, it should be assumed that initially the information will spread among those recipients who won't be attached to the product in the end. Here the target group develops in a dynamic process, which is a result of the detachment of individuals from the disseminating group. This process of detachment is what will define the group afterwards.
In this way, there are quite a few risks for the product. For this reason, besides the design of the information, it is important to choose the right
seeders. The seeders receive the information first. Their fundamental approach towards the product will be crucial for how the information will be received further. However, one should not over-estimate this influence and not be afraid of negative perceptions as well. In each generation, the perception will be modified and several different perspectives will average in the end - especially in the case of IDI. This progress should be considered as acceptable and as not to be avoided. In the end, the target group will anyways be the one, that holds a positive view on the story and therefore, the product.

(Note: The aspects of resentment decrementation and diffusion blockers are not discussed in this essay, due to its large scope of necessary explanation. These aspects - the Resentment and Diffusion Theory - are also not part of the original thesis, since they were developed after the degree course.)

[Ill. 6.: IDI model]

The effectiveness of the story can be rated in a supplementary controlling. Thereby, the quantitative dissemination (as far as measurable), the size of the detached target group as well as the effectiveness of the catenation with the product are crucial. A failure would be, for instance, that the story hardly got disseminated, just a small target group got detached or that the link to product got lost during the progress. An epidemic but negative view on the product however, cannot be interpreted as a failure. (Not in the sense of marketing - maybe in the sense of the product design.) A success would be, that the expectation of a large dissemination was fulfilled, which led to a large target group which clearly can identify the product and has a positive view on it. In both cases the model helps to analyse, e.g. which parameters could be modified for further actions or which should remain unaltered.

5. Conclusion

As a result of the study it can be concluded, that the actual win of the research is: The developed perspective which specifically allows to follow a strict and logical model of operation. The strength of such a model is, that there is a reason for a decision to do this or that. Therefore in the retrospective analysis of e.g. a marketing action, there is a structure which allows the relative classification of the results (positive or negative). That is the only way to learn from experience and to avoid mistakes in future actions. Actions which are not based on a logical model are similar to wild guessing. And the probability to repeat the same mistake the next time is very high.
The theories and models discussed in this essay are not yet completed. Every action and evaluation will modify them. Now and in the end, there is and will be a model which functions for a specific time for a specific milieu in a specific culture area. This is what media science predicts. And what professional practice demands.

6. Reference list

Arndt, J. (1967). Role of Product-Related Conversations in the Diffusion of a New Product. Journal of Marketing Research, 4 (3), 291-295.

Bernays, E. (2014). Propaganda - Die Kunst der Public Relations. 5. Auflage (1. DE). Freiburg: orange press.

Le Bon, G. (1964). Psychologie der Massen. Stuttgart: Alfred Kröner Verlag.

Brednich, R. (1990). Die Spinne in der Yucca-Palme. München: Verlag C. H. Beck.

Brednich, R. (1993). Das Huhn mit dem Gipsbein. München: Verlag C. H. Beck.

Brednich, R. (1994). Die Maus im Jumbo-Jet. München: Verlag C. H. Beck.

Brednich, R. (1996). Die Ratte am Strohhalm. München: Verlag C. H. Beck.

Brednich, R. (2004). Pinguine in Rückenlage. München: Verlag C. H. Beck.

Broch, H. (1979). Massenwahntheorie. 1. Auflage. Frankfurt a.M.: Suhrkamp.

Buber, M. (1973). Das dialogische Prinzip. Heidelberg: Verlag Lambert Schneider GmbH.

Canetti, E. (1980). Masse und Macht. München: Fischer Taschenbuch Verlag.

Cnyrim, P. (2015). Slenderman und Smile Dog. 1. Auflage. München: riva Verlag.

Czerwinski, W. (2014). Filmmarketing im Social Web. Saarbrücken: AV Akademikerverlag.

Darwin, C. (1872). Der Ausdruck der Gemütsbewegungen bei dem Menschen und den Tieren. Frankfurt a.M.: Eichborn Verlag.

Deleuze, G. (1989). Das Bewegungs-Bild. 1. Auflage. Frankfurt a. M.: Suhrkamp.

Deleuze, G. (1997). Das Zeit-Bild. 1. Auflage. Frankfurt a. M.: Suhrkamp.

Eilert, D. (2013). Mimikresonanz: Gefühle sehen. Menschen verstehen. Paderborn: Junfermann.

Ekman, P. (1991). Telling Lies. New York: W. W. Norton.

Ekman, P. (2007). Emotions revealed: Recognizing faces and feelings to improve communication and emotional life. New York: Henry Holt.

Ellgring, H. (1989). Alles über Körpersprache. 12. Auflage. München: Wilhelm Goldmann Verlag.

Freud, S. (1974). Massenpsychologie und Ich-Analyse. Frankfurt a. M.: Fischer Taschenbuch Verlag.

Frenzel, K. (2008). Storytelling - Das Harun-al-Raschid-Prinzip. München/Wien: Hanser.

Glaser, B.und Strauss A. (1998). Grounded Theory: Strategien qualitativer Forschung. Göttingen: H. Huber.

Hartung, J. (2000). Sozialpsychologie. Stuttgart: Kohlhammer GmbH.

Horn, E. (2009). Schwärme - Kollektive ohne Zentrum. Bielefeld: transcript Verlag.

Jefferson, P. und Socolow, M. (2013). The Myth of the War of the Worlds Panic. [online] slate.com. Verfügbar unter: http://www.slate.com/articles/arts/history/2013/10/orson_welles_war_of_the_worlds_panic_myth_the_infamous_radio_broadcast_did.single.html [Aufgerufen am 12. April 2016].

Jüngst, H. (Hrsg.). (1999). Urban Legends. Stuttgart: Philipp Reclam jun..

Kirby, J. (Editor). (2006). connected marketing - the viral, buzz and word of mouth revolution. 1. Auflage. Burlington (US): Elsevier Ltd..

Kuhn, T. (1978). Die Entstehung des Neuen. 1. Auflage. Frankfurt a. M.: Suhrkamp.

Kumbier, D. (2014). Interkulturelle Kommuniation: Methoden, Modelle, Beispiele. 7. Auflage. Hamburg: Rowohlt.

Kuokkanen, J. (1997). A Link Between Consumer Dis/Satisfaction and Postpurchase Word-of-Mouth Intentions: An Experimental Study. Turku: Turku School of Economics and Business Administration.

Krüll, C. (2012). Körpersprache - Wahrnehmen, erkennen, deuten. Augsburg: C. H. Beck.

Lachmann, R., Butterfield, E. und Lachmann, J. (1979). Cognitive Psychology and Information Processing: And Introduction. Philadelphia (PA): Taylor & Francis.

Lange, G. (Hrsg.). (2003). Moderne Sagen. Stuttgart: Philipp Reclam jun..

Lutter, M. (2012). Soziale Strukturen des Erfolgs: Winner-take-all-Prozesse in der Kreativwirtschaft. Köln: Max-Planck-Institut für Gesellschaftsforschung.

Lis, B. und Korchmar S. (2013). Digitales Empfehlungsmarketing - Konzeption, Theorien und Determinanten zur Glaubwürdigkeit des Electronic Word-of-Mouth. Wiesbaden: Springer Gabler.

Molcho, S. (1983). Körpersprache. 23. Auflage. München: Wilhelm Goldmann Verlag.

Molcho, S. (2001). Alles über Körpersprache. 12. Auflage. München: Wilhelm Goldmann Verlag.

Möller, M. (2011). Online-Kommunikationsverhalten von Multiplikatoren. Wiesbaden: Gabler Verlag | Springer Fachmedien.

Navarro, J. (2008). Menschen lesen. 8. Auflage. München: Münchener Verlagsgruppe.

Navarro, J. (2013). Menschen verstehen und lenken. 5. Auflage. München: Münchener Verlagsgruppe.

Nießing, D. (2007). Kunden-werben-Kunden-Kampagnen. Wiesbaden: Deutscher Universitäts-Verlag.

Oetting, M. (2009). Ripple Effect - How Empowered Involvement Drives Word of Mouth. Wiesbaden: Gabler | GWV Fachverlage GmbH.

Podsiadlowski, A. (2004). Interkulturelle Kommunikation und Zusammenarbeit. München: Verlag Franz Vahlen GmbH.

Pütz, C. (2009). Wirkung von electronic-Word-of-Mouth-Empfehlungen bei misstrauischen Rezipienten. Dissertation. Rheinisch-Westfälische Technische Hochule Aachen / Fakultät für Wirtschaftswissenschaften.

Reichelt, J. (2013). Informationssuche und Online Word-of-Mouth. Wiesbaden: Springer Gabler.

Rothe, F. (2006). Zwischenmenschliche Kommunikation. Wiesbaden: Deutscher Universitäts-Verlag.

Röthlingshöfer, B. (2008). Mundpropaganda-Marketing. München: Deutscher Taschenbuch Verlag.

Salber, W. (2006). Goethe zum Film - Morphologische Markt- und Medienpsychologie. Bonn: Bouvier Verlag.

Sammer, P. (2014). Storytelling - Die Zukunft von PR und Marketing. Köln: O'Reilly.

Secord, P. und Backman C. (1977). Sozialpsychologie. 2. Auflage. Frankfurt a. M.: Fachbuchhandlung für Psychologie / Verlagsabteilung.

Scholz, A. (2009). Filmmarketing 2.0. Marburg: Tectum Verlag.

Schüller, A. (2015). Das neue Empfehlungsmarketing. Göttingen: BusinessVillage GmbH.

Skinner, B. (1953). Science and Human Behavior. Upper Saddle River (NJ): Pearson Education.

Tadié, J. und M. (2004). Im Gedächtnispalast: Eine Kulturgeschichte des Denkens. 2. Auflage. Stuttgart:Klett-Cotta.

Tercia, C. (2010). Consumer Intention to Participate in Mobile Based Word-of-Mouth Activities: The Role of Incentive and Tie-Strength from Senders and Receivers Perspectives. Dissertation. Universität Hamburg / Fakultät Wirtschafts- und Sozialwissenschaften.

Trommsdorff, V. (2002). Konsumentenverhalten. 4. Auflage. Stuttgart: Kohlhammer.

Vogelaar, R. (2011). Der Superpromoter. 1. Auflage. Wiesbaden: Gabler Verlag | Springer Fachmedien.

Waschkau, A. und Waschkau A. (2014). The Hoax-Files. Hannover: jmb-Verlag.

Watzlawick, P. (1988). Wie wirklich ist die Wirklichkeit. 16. Auflage. München: R. Piper & Co Verlag.

Watzlawick, P. (2011). Menschliche Kommunikation. 12. unveränderte Auflage. Bern: Verlag Hans Huber.


This document is presented by openpot Media | copyright Benjamin Czycholl 2016 | last modified on 2016.06.10 | the author reserves all rights to texts and illustrations in this document | for research quotation please refer to the original document