Running head Past tense processing Address for correspondence

Evaluating Behavioral and Neuroimaging Data

on Past Tense Processing

Mark S. Seidenberg

James H. Hoeffner

Neuroscience Program and

Departments of Psychology and Linguistics

University of Southern California

Running head: Past tense processing

Address for correspondence:

Mark S. Seidenberg

Neuroscience Program

University of Southern California

Los Angeles, CA 90089-2520

email: marks@https://www.wendangku.net/doc/3c2802040.html,

fax: 213-740-5687

phone: 213-740-9174

Abstract*

Jaeger, Lockwood, Kemmerer, Van Valin, Murphy, and Khalak (1996) reported an experimental study that provided reaction time and PET neuroimaging data said to support Pinker’s (1991) theory of inflectional morphology in which rule-governed forms and exceptions are processed by separate mechanisms. The results were also taken as evidence against connectionist accounts in which a single processing system generates both types of forms. We provide a critical analysis of the study that yields three main conclusions: First, Jaeger et al.'s data do not provide strong evidence that rule-governed forms and exceptions are processed in separate brain regions. Second, there are problems with the design of the study that contaminate critical comparisons between conditions. The results therefore afford alternative interpretations related to experiment-specific factors rather than the regular-irregular distinction. Third, the dissociations between rule-governed forms and exceptions observed in studies such as Jaeger et al.'s can be accommodated by the connectionist theory. We conclude by offering suggestions for future research that would overcome the major limitations of this study and provide more decisive evidence bearing on the issues.

* Research supported by National Institute of Mental Health grants PO1-47566 and KO2-01188. The manuscript benefitted from the comments of the editor and three anonymous reviewers, whom we thank. We also thank Joe Allen (USC) and Jay McClelland (CMU) for helpful discussions. Jim Hoeffner is now at the department of psychology, University of Memphis.

1. Introduction. The formation of the past tense of verbs in English has been the focus of a long-running debate about the relevance of connectionist concepts to understanding language. 'Connectionism' refers to a body of ideas about knowledge representation, acquisition, and processing that have been applied to a broad range of issues concerning language and other aspects of cognition (Rumelhart & McClelland 1986). Connectionist theories attempt to explain phenomena in terms of networks of simple, neuron-like processing units whose behavior changes adaptively through learning. Such theories are typically instantiated as computational models that simulate detailed aspects of human performance (e.g., Seidenberg & McClelland 1989). Part of the attraction of the approach is that the same concepts are also being used in developing theories of the brain mechanisms underlying behavior (e.g.,Gluck & Rumelhart 1990; Elman et al. 1996), offering the possibility of a unified account of the two.

In one of the early studies within this approach, Rumelhart and McClelland (1986) implemented a model of some aspects of the acquisition of the English past tense. The model was a past tense generating device that took the present tense of a verb as input and produced the past tense as output. Its main interest lay in the fact that it could generalize to novel forms (e.g., wug-wugged). The model was therefore said to provide an alternative to approaches in which generativity results from the use of rules. Several limitations of this model were subsequently noted by Pinker and Prince (1988). These limitations were said to vitiate Rumelhart and McClelland's claim to have provided an alternative account of linguistic knowledge, and were used to call into question the relevance of connectionist concepts to understanding language: Pinker and his colleagues subsequently acquired a large amount of evidence from a variety of domains said to support a 'modified traditional theory' of the past tense employing two mechanisms: a rule used to generate regular forms (e.g., walk-walked and perform generalization (e.g., wug-wugged), and an 'associative net' used in generating irregular forms (e.g., take-took) (see Pinker 1991 for a summary). Connectionists have also presented modified models that addressed many of the limitations of Rumelhart and McClelland's original work (see, for

example, Daugherty & Seidenberg 1992; MacWhinney & Leinbach 1991) and have challenged findings taken as supporting the dual-mechanism theory (e.g., Hoeffner, 1997).

Because the past tense of English is a minor aspect of language and because this controversy has continued without resolution for almost 10 years, it is worth remembering why these issues are important. Studies of the past tense (and inflectional morphology more generally) have historically played important roles in several core areas of linguistic research, including phonological theory (e.g., Halle & Mohanan 1985), language acquisition (e.g., Berko 1958; Bybee & Slobin 1982), adult psycholinguistics (e.g, Bybee & Moder 1983), diachronic change (e.g., Hare & Elman 1995), and the genetic (Rice 1996) and neurobiological (Ullman et al. 1997) bases of language. These studies illustrate the value of exploring a specific aspect of language in considerable detail. However, it is also true that if the connectionist approach were proved to be incapable of explaining something as simple as the past tense, this result would impose severe limitations on its relevance to other, more complex aspects of language.

The article by Jaeger, Lockwood, Kemmerer, Van Valin, Murphy and Khalak (1997; hereafter J et al.) seems to provide compelling evidence bearing on this controversy. Pinker's theory draws a strong distinction between rule-governed forms and exceptions: they are produced by separate subsystems that involve different types of knowledge representations (a rule, an 'associative net'), learning procedures (rule induction, rote learning), and processing mechanisms (rule application, lexical look-up). In the connectionist approach, a single computational system is used to generate all forms. The same principles of knowledge representation, learning, and processing therefore govern both types of forms. The essential idea is that a network trained to generate past tenses can find a set of weights that allow it to produce all forms. The network therefore does not embody a categorical distinction between 'rule-governed' and 'exception.' Using a single mechanism to generate all forms is intended to capture the fact that rule-governed forms and exceptions are not categorically different; they share structure in at least three respects. First, there are systematic correspondences between the present and past tense forms of both regular and irregular verbs. For example, in both bake-baked and take-took, the past tense retains

the onset and coda of the present tense. The degree of overlap varies across verbs but there is some for all past tense alternations except the suppletive ones. Second, there are similarities between regular and irregular past tense forms. For example, crept and slept, which are nominally irregular, end in /t/, which is the regular past tense inflection in cropped, stepped and many other regular forms. Finally, there are subregularities among pools of irregular verbs, such as sing-sang/ring-rang, and blow-blew/throw-threw/grow-grew. These properties motivated earlier treatments such as Halle and Mohanan (1985) and Bybee (1985). The neural network approach shows how these partial regularities can be picked up in the course of training and encoded by the weights along with the regularities characteristic of the past tense rule. The weights that are derived from exposure to examples can also be used to generate past tenses for novel forms such as wug.1

J et al. presented the results of a positron emission tomography (PET) neuroimaging study providing evidence about brain activity during past tense generation. The data were interpreted as indicating that rule-governed and irregular forms activate different brain regions, consistent with Pinker's dual-mechanism theory and contradicting the connectionist approach. Data concerning subjects' latencies to produce past tenses provided additional support for this conclusion.

The main point of this article is that there are conceptual, methodological, and interpretive flaws in the J et al. study that undermine their claims. Although we agree that neuroimaging studies have the potential to provide important evidence bearing on cognitive and linguistic theories, this particular study does not succeed in this regard. We will discuss three main points: First, J et al.'s data do not provide strong evidence that rule-governed forms and exceptions are processed in separate brain regions. Second, the design of the study introduced confounding factors that contaminate critical comparisons between conditions. These affect both the imaging and latency data. The results therefore afford alternative interpretations related to experiment-specific factors rather than the regular-irregular distinction. Third, the assumption that dissociations between regular and irregular forms both implicate separate mechanisms and contradict the connectionist theory needs to be examined closely. The particular dissociations that

J et al. were seeking are compatible with both types of theories and so do not discriminate between them. The same issue arises in connection with other recent reports of such dissociations (Marslen-Wilson & Tyler 1997; Ullman et al 1997). In closing we suggest directions for future research, describing alternative hypotheses that would distinguish between the theories and methodologies appropriate for addressing them.

2. The PET Data. J et al.'s study provides data concerning brain activity while subjects were generating the past tenses for known, rule-governed forms (e.g., pull-pulled), irregular forms (sweep-swept) and nonce forms (baff-baffed).2 Stimuli were presented in separate blocks of 46 items of each type, yielding the Regular, Irregular, and Nonce Past conditions, respectively. The PET images reflected average activity during a block of 46 trials of a single type. In two control conditions, subjects simply read words aloud rather than generating past tenses. The Read Verb control condition consisted of a mixture of regular and irregular verbs; in the Read Nonce control condition, subjects read aloud a different set of nonce verbs. Brain areas associated with generating past tenses for regular verbs were determined by subtracting the pattern in the Read Verb condition from the pattern in the Regular Past condition; analogous subtractions were performed for the Irregular and Nonce conditions. J et al. concluded that: 'there are major differences in both the location and the amount of brain activation in the regular vs. irregular tasks; specifically, overlapping yet partially distinct sets of brain areas are recruited for the computation of regular vs. irregular forms, with a much larger area of activation involved in the irregular computation. Thus...our data are consistent with the dual-systems theories and are problematic for the single-systems accounts.' (p. 484)

Table 1 presents the data in J et al.'s Table 4 (p. 467) but regrouped to make it easier to see the results for each verb condition. Three aspects of these data (highlighted in bold) were taken as providing crucial support for the dual-mechanism theory. First, part of the left dorsolateral prefrontal cortex was active in the Regular Past and Nonce Past conditions, but not the Irregular Past condition (after the relevant control conditions were subtracted out). This area was thought to

be involved in generation by rule. Second, the left mid-temporal gyrus was active in the Irregular Past condition but not in the Regular Past or Nonce Past conditions (against subtracting out the relevant controls). This area was said to reflect retrieval of stored irregular forms. Finally, the left lateral orbito-frontal cortex was active in both the Irregular Past and Nonce Past conditions. J et al. argue that this area is involved in 'inhibiting default, high frequency responses, and is inactive during overlearned, practiced behaviors' (p. 483), the latter including generating a known rule-governed form.

________________________________________

Insert Table 1 About Here

________________________________________

This interpretation of the data raises two questions: First, what pattern of data would count as evidence against the dual-mechanism theory, and second, what is the basis for the attributions about what is being computed in different brain regions? The dual-mechanism theory suggests that regular and nonce verbs should pattern together and differ from irregular verbs. The first two findings mentioned above are consistent with this prediction: there were regions activated for rule-governed forms but not exceptions and vice versa. However, by the same logic we would not expect to see areas that are activated for both regular and irregular verbs or for irregular and nonce verbs, yet the data also yielded these patterns. In fact, there were unique brain regions activated for all three pairwise combinations of conditions, and there were areas activated by all three types of stimuli as well. Given the absence of any independent theory about why particular regions might be involved in processing particular types of stimuli, this data pattern cannot be taken as providing strong evidence for the dual-mechanism theory. The fact that some areas were activated for both irregulars and nonce items or for regulars and irregulars could just as well be taken as decisive evidence against the theory. In fact there are more areas activated in common by the irregulars and nonce words, which are supposed to be processed by separate mechanisms, than the regulars and nonce words, which are supposed to be generated by the rule.

J et al. discuss their data fully and attempt to provide interpretations of the activity in each brain region consistent with the dual-mechanism theory. However, their account turns on

attributions about the functions associated with different brain regions that are highly questionable; they lack independent motivation and they are not internally consistent. For example, in explaining the activity in the left dorsolateral prefrontal cortex associated with the regular and nonce conditions, J et al. argue that this area of the brain is responsible for ‘using knowledge stored in long-term or short-term memory to formulate novel conceptual and behavioral sets in a willful, internally generated manner, as opposed to a purely stimulus-driven manner.’ (p. 482). They also state that this area of the brain has been shown to be involved in the ‘on-line formulation of intentional novel behaviors.’ (p. 488). This account suggests that applying linguistic rules is ‘willful’ rather than merely ‘stimulus-driven’ but this characterization of the use of linguistic knowledge conflicts with the views of most psycholinguists, who emphasize the rapid, automatic, unconscious nature of the processes involved in comprehension and production. Moreover, whereas generating past tenses for nonsense words may be ‘novel’behavior, generating past tenses for known regular verbs is not. To the contrary, it is a highly practiced behavior. Why a brain area would be activated for behaviors that are willful and novel (generating past tenses for nonce words) and familiar and unconscious (generating known past tenses for familiar words) is unclear and unsupported by any independent neurophysiological evidence. Below we offer an alternative account of the activity of this region that is motivated by independent evidence and helps to explain other aspects of J et al.'s data.

J et al.'s interpretation of the activity in the left dorsolateral prefrontal cortex creates other problems when it comes to explaining the finding that the irregular and nonce conditions activated a common region (the left superior frontal gyrus). J et al. claim that this area is involved in ‘inhibiting default, high-frequency responses, and is inactive during overlearned, practiced behaviors.’ (p. 482). This interpretation assumes that producing the regular inflection is a ‘default, high-frequency behavior’ that is an ‘overlearned, automatic response’ (p. 487) and must be actively suppressed in producing the irregular. However, the interpretation of the activity in the left dorsolateral prefrontal cortex discussed above assumed that generating the regular suffix was ‘intentional’ and ‘willful’ behavior. Moreover, although it is clear why the regular past tense

mechanism might have to be suppressed in generating an irregular past tense, this reasoning does not extend to the nonce task, where the regular suffixation process needs to apply, not be suppressed. J et al. address this apparent contradiction by arguing that in the nonce task this area is not inhibiting the regular suffixation process per se but is "involved in inhibiting any number of possible irrelevant responses to this novel list of forms." (p. 487). The problem with this argument, aside from its post hoc nature, is that this brain area is now assumed to serve different functions in the processing of irregulars and nonce words. Thus, when the regular and nonce verbs activated a common brain region, it was taken as evidence for a single mechanism underlying the two but when the irregulars and nonce items activated a common area it was interpreted in terms of two mechanisms. This inconsistency illustrates the weak linkage between theory and data in J et al.'s account.

J et al.'s assumptions about localization of function should also be considered in light of two recent studies of brain-injured patients' impairments in verb morphology (Ullman et al. 1997; Marslen-Wilson & Tyler 1997). All three studies claimed to find support for the dual mechanism theory, but they yielded conflicting evidence with respect to where the putative subsystems are localized. Data from several types of brain injured patients led Ullman et al. to conclude that a frontal lobe/basal ganglia system is responsible for regular inflection, with irregulars processed by a separate system located in temporal-parietal cortex. However, J et al.'s data implicate various parts of the frontal lobes in both regular and irregular inflection. In the Generate Past-Read Verb comparisons, four areas in the left hemisphere’s frontal lobes were active for regular verbs, but three partially overlapping areas of the frontal lobes were also active for the irregular verbs. Similarly, parts of the temporal and parietal lobes were active for irregulars, as Ullman et al. predict, but also regular and nonce verbs. Finally, J et al. did not observe above threshold activation of the basal ganglia in any condition. In Marslen-Wilson and Tyler’s study, it was damage to the right hemisphere that played an important role in the impairment of the irregular system but neither Ullman et al nor J et al. ascribe a critical role to this region. In addition, the patients with impaired performance on regular items in Marslen-Wilson & Tyler’s study had

damage to left hemisphere temporal and parietal lobes, which according to Ullman et al. should have led to impaired irregular performance. Our view is that these inconsistencies are due in part to the fact that the regular-irregular distinction does not capture the relevant generalization about performance, which is affected by the frequency and consistency of the mapping between present and past tenses. In any case, the complex and partially conflicting patterns of results in these studies suggest that considerable additional research is needed in order to understand the brain bases of past tense representation and processing.

3. Methodological issues. Even if J et al.'s data had yielded clearly different patterns for rule-governed forms and exceptions, they could not be taken as decisive because of a serious flaw in the design of the study. The flaw relates to the presentation of different types of stimuli in homogeneous blocks. Each of the critical conditions (Regular Past, Irregular Past, Nonce Past) consisted of 46 items of that type. This blocked design is imposed by the low temporal resolution of the PET methodology, which requires subjects perform the same type of task over the interval during which the image is computed. The presentation of stimuli in homogeneous blocks introduces confounding factors that invalidate the comparisons between conditions.

In the Regular Past condition, subjects had to add the regular suffix 46 consecutive times. All that was required was determining which of the three allomorphs of the regular suffix (illustrated by jump-jumped, love-loved, pat-patted) was correct for each verb. These allomorphs are phonetically similar, differing only in the presence or absence of voicing and/or the insertion of an unstressed vowel. Choice of allomorph is determined by the final phoneme of the present tense verb. The 46 items in the Irregular Past condition differed in several respects. They form the past tenses in a wide variety of ways including several vowel changes (e.g., fall-fell; sit-sat; bite-bit), changing final /d/ to /t/ (e.g., build-built; send-sent), and both vowel and consonant changes (bring-brought; keep-kept; stand-stood; sell-sold). They overlap in more complex ways than the regulars: J et al.'s stimuli included verbs that rhyme and form their past tenses similarly (e.g., tell-told; sell-sold) but also verbs that rhyme and form their past tenses differently (e.g., bring-

brought; sing-sang). There were also partially-overlapping trials such as feed-fed vs. sweep-swept (same vowel change, different treatment of final consonant) and some irregular present tenses rhyme with the past tenses of other irregulars in the list (e.g., hold rhymes with told and sold). Finally, generating the correct irregular past tense requires recognizing the present tense verb as a particular lexical item, not just its final phoneme.

Given these differences between regular and irregular verbs, presenting the stimuli in blocks introduced confounding factors that are certain to have affected performance. One confound is that the blocks created different kinds of priming effects. A large body of psychological research has addressed how performance on one trial in an experiment is affected by performance on similar or dissimilar preceding trials. Such priming effects are ubiquitous; they occur for adjacent items (e.g., Forster & Davis 1984) and for items separated by many intervening trials (e.g., Scarborough et al. 1977). Stimuli that are similar create facilitation; stimuli that are dissimilar create interference. To take a simple example, Taraban and McClelland (1987) showed that pronunciation of the word gave is facilitated if it is preceded by the rhyming word save but inhibited by have. Unless they are themselves the focus of study (as in Gordon, 1983), experimenters typically take pains to avoid repeating similar or identical stimuli within an experiment. J et al. could not do this, however, because of limitations of the PET methodology. Presenting stimuli in homogeneous blocks creates massive priming effects that differ across conditions. Whereas the high degree of similarity across trials in the Regular Past and Nonce conditions facilitated performance (as indicated by the latency data discussed below), the variability across trials in the Irregular Past creates interference. Readers can verify this by conducting the experiment themselves. Table 2 contains the first 23 irregular stems used by J et al. in the same order as they were presented to subjects. Read through the list generating the past tenses quickly and accurately. There should be a sense that forming the past tense differently on each trial creates interference. Table 2 also presents a list of regular verbs for comparison. Table 3 presents a list of irregular verbs grouped into families that pattern similarly. This grouping simplifies the task considerably. Note that J et al.'s Regular Past condition created even stronger

grouping effects because all 46 items conformed to the single regular pattern.3 Thus, the blocks of regular and irregular stimuli created different priming effects due to differences in the amount and degree of overlap across trials that are confounded with type of verb. Studies of priming effects using PET and functional magnetic resonance imaging (fMRI) have shown that different kinds of priming produce different patterns of brain activity (e.g., Gabrieli et al, 1996; Demb et al., 1995). Given J et al.'s design, it cannot be determined whether their PET images reflect different types of priming or verbs.

________________________________________

Insert Tables 2 and 3 About Here

________________________________________

A second consequence of presenting the stimuli in blocks is that the Regular and Nonce Past conditions differed from the Irregular Past condition in terms of how deeply stimuli had to be processed in order to generate correct responses. In the Regular and Nonce conditions, subjects knew in advance of each stimulus that the regular past tense suffix would be required. The form of the past tense was determined by phonological properties of the present tense verb. In the Irregular Past condition, however, subjects did not know in advance how the past tense would be formed; producing correct responses required identifying each present tense verb as a particular lexical item.The Regular and Nonce Past conditions therefore differed from the Irregular Past condition in terms of the depth of processing (Craik & Lockhart 1972) that was required: the Irregular condition required a deeper analysis of each item than the other conditions. This confound is important because other neuroimaging studies indicate that stimuli that are processed more deeply produce different brain activity than stimuli that are processed more superficially (e.g., Gabrieli et al. 1996). Note that in a design in which regular and irregular forms are randomly intermixed this confound is eliminated: if the subject does not know in advance whether the next stimulus will be regular or irregular verb, all verbs have to be processed equally deeply.

J et al. acknowledged that their subjects quickly realized that all the verbs in Regular Past condition were regular (p. 476) and expressed appropriate concern that this aspect of the study might have affected the results, noting that 'The obvious test of the influence of task format is to

gather RT’s to stimuli where an equal number of frequency-matched regular and irregular stems are randomly mixed.' (p. 476). As they observe, PET methodology is incompatible with such a design because it requires subjects to be exposed to many consecutive examples of a given type. Unfortunately, all this indicates is that PET is an inappropriate method for addressing the issues at hand. Acknowledging that there are confounds between the conditions that are likely to influence the results does not eliminate the problem. A better alternative would have been to consider using methods that do not impose the same restrictions as PET, one of which we discuss below.4

4. Anomalies in the latency data. Our assertions about the effects of blocking the stimuli are supported by J et al.'s latency data and help to explain some anomalous aspects of them. J et al.'s data can be compared to that from a similar study conducted by Seidenberg and Bruck (1990, hereafter SB; Seidenberg 1992).5 In SB's study (see Table 4), subjects were presented with 40 regular and 40 irregular present tense verbs (equated in frequency) and asked to generate the corresponding past tense forms. The design was the same as J et al's except that the regular and irregular verbs were intermixed rather than presented in separate blocks. Comparisons between the two studies therefore provide direct evidence about the effects of blocked vs. mixed presentation.

________________________________________

Insert Table 4 About Here

________________________________________

Several aspects of the data indicate that presenting the stimuli in blocks had a huge impact on the results. J et al.'s subjects were much faster than SB's and the speed-up was largely due to the regular verbs, which yielded mean latencies of 524 ms and 925 msec in the J et al. and SB studies, respectively. This pattern is what would be expected if presenting stimuli in blocks allowed subjects to take advantage of prior knowledge that all the stimuli were going to be regularly inflected. The subjects in SB’s study had net generation latencies (latency to generate past tense - latency to read verb aloud) of 350 ms for regular verbs and 455 ms for irregular verbs. In contrast, the mean generation latencies in these conditions in J et al.'s study were 38 ms

and 238 ms. Thus, regular past tenses were produced approximately 10 times faster in the J et al. experiment. There was also a smaller speedup for the irregular verbs. These data strongly indicate that the subjects in the J et al. study were using response strategies conditioned on the presentation of stimuli in blocks. The fact that it only took 38 ms on average to generate regular past tenses is itself striking: other research suggests that it takes at least 100 msec to merely identify a word and fixation durations for words in texts typically average 200-250 msec (Rayner & Pollatsek 1994). It is hard to imagine that applying a morphological rule would take an order of magnitude less time to perform than these tasks unless subjects could exploit the knowledge that all items in the block were regular.

The data in J et al.'s nonce conditions are also inconsistent with the results of other studies. In most studies, generating responses for familiar items is much easier than generalizing to novel forms. For example, subjects are faster at generating the pronunciation of a familiar word such as must than a nonword like nust. In the J et al. study, however, subjects produced past tenses for nonsense words slightly faster than for familiar irregular words. Table 4 presents the data from the J et al. study and a similar experiment by Seidenberg and Bruck (1990).6 Stimuli were again blocked in the J et al. study; in the SB study, they were intermixed with an equal number of known regular and irregular verbs. The SB data are means from 52 nonce items and 30 subjects. Subjects in the two studies took roughly the same amount of time to name the nonwords, but SB's subjects took far longer to generate their past tenses. Latencies in the SB study were longer because subjects did not know that the stimulus on every trial would end with the regular past tense. Subjects were simply told that the stimuli would include both familiar and unfamiliar forms and were instructed to generate past tenses for the unfamiliar ones 'as if' they were real verbs in the language.

The comparisons between these studies clearly point to large effects attributable to demand characteristics of the J et al. experiment. Presenting the stimuli in homogeneous blocks provides information that subjects can use in planning their responses. When the stimuli are all regular, subjects can respond knowing that all that is required is adding the correct phonologically-

conditioned regular suffix. The task is highly predictable and repetitious, yielding latencies that are little longer than those for simply saying the verbs aloud. Because the identity of the stimulus is not critical to producing the response, it matters little whether it is a familiar word or a nonword. Blocking the irregulars is also helpful insofar as subjects know that the regular past tense will not be required. However, the benefit is smaller than in the regular case because subjects must still process each verb deeply enough to determine its past tense; moreover, forming the past tense differently on each trial produces interference that slows performance. These effects are experiment-specific and have little relevance to normal performance, in which speakers and listeners routinely switch between regular and irregular pasts and other tenses.

To summarize, effects due to repetitions of stimuli and other nonrandom aspects of stimulus presentation are robust, well-studied phenomena in experimental psychology. They are so prominent that in studies employing standard reaction time measures researchers use experimental designs that avoid them or allow their effects to be assessed. There are literally dozens of experiments in the literature examining the processing of different types of words (e.g., abstract-concrete; homophone-nonhomophone; high-low frequency; familiar-unfamiliar; regular-irregular pronunciation; word-nonword) and the stimuli are never blocked by type unless the explicit purpose of the experiment is to examine blocking itself. The reasons why researchers avoid this design feature are illustrated by J et al.'s study: repeating stimuli of a given type creates different kinds of priming effects across conditions and encourages condition-specific strategies.

In presenting stimuli in homogeneous blocks, J et al. were following the standard PET methodology pioneered by Raichle and his colleagues (e.g., Petersen, Fox, Posner, Mintun, & Raichle, 1988). (199x). Every methodology has its limitations; in the case of PET, it is the slow temporal resolution of the measure, which necessitates averaging across a block of similar trials. The use of PET to localize cognitive functions has generated considerable controversy about basic aspects of the methodology (Roskies 1994), including what brain events PET actually measures (Barinaga, 1997), the validity of the subtractive methodology (Sergent 1994), and the reliability of localization claims (Poeppel 1996). As with other methodologies, the utility of the approach

depends on the kinds of questions being addressed. Raichle and his colleagues' classic research, for example, involved comparisons between conditions such as passively looking at actual words vs. nonsense words or reading verbs vs. generating an associated action. In these cases, each trial carries very little information about what will occur on other trials and subjects cannot formulate response strategies that work for one condition but not another. Hence the ceteris paribus condition that is the foundation of experimental research was largely met and the studies yielded coherent findings. In J et al.'s study, however, presenting the stimuli in homogeneous blocks introduced several confounding factors that invalidate the attribution of the observed effects to the nominal independent variable, type of verb.

5. An alternative interpretation: more difficult = more activation. Once the consequences of presenting stimuli in blocks are recognized, it is apparent that J et al.'s data afford a completely different interpretation that is consistent with the single mechanism theory. Because of the way the stimuli were selected and presented, the regular, irregular and nonce conditions differed in difficulty. The regular verbs were easiest, the nonce verbs were slightly harder, and the irregular verbs were much harder. A simple interpretation is that the observed patterns of brain activity reflected the relative difficulty of these conditions. The data are therefore consistent with the single-mechanism theory in which performance is determined by the degree of difficulty in the mapping between present and past tense, not type of verb.

The fact that there is a confound between type of verb and level of difficulty is crucial to understanding the PET data and claims about localization of function. J et al.'s data indicate that the irregulars produced the longest generation latencies and the most errors, the regulars yielded the fastest responses and fewest errors and the nonce verbs fell in between. These differences in difficulty are not intrinsic to these types of stimuli; it would be possible to equate the conditions in terms of difficulty by using different items (e.g., harder regular verbs or easier irregulars) and that is necessary to allow valid comparisons between conditions. Research by Just et al. (1996) has indicated that task difficulty is correlated with amount of brain activity. Just et al. conducted an

fMRI study of language comprehension that involved three main conditions in which subjects read sentences varying in syntactic complexity. The relative difficulty of the stimuli was independently established from reaction time data. The main finding was that the difficulty of the materials was correlated with amount of brain activity: the harder stimuli produced activation over a broader area than the easier stimuli.

J et al.'s study replicates this result: there was a high correlation (.835) between the difficulty of the conditions (as indicated by response latencies) and the amount of brain activity they elicited. The irregular condition activated the most areas (15, 3 of which reached significance by the strictest thresholding method) and a total of 2,912 pixels were active after the baseline condition was subtracted. The nonce condition produced activation across 14 areas (none of which reached significance by the strictest criterion) and 1,368 pixels. The regular condition produced the least activation, only 728 active pixels in 7 areas, none of which reached significance by either the strictest or second-strictest criteria.

Consider now the three findings that J et al emphasize in their discussion of the Generate Past-Read Verb subtractions. One area, the left mid-temporal gyrus (area 21), was active for irregulars but not for nonce or regulars. J et al.'s interpretation is that this area was responsible for irregular inflection. However, this activity may simply reflect the fact that the irregular condition was the most difficult, and more difficult conditions produce more brain activity over a broader area. The inference that the area was involved in processing irregular verbs would only be valid if the conditions had been equated in terms of difficulty.

The second area J et al. emphasized is the left superior frontal gyrus (area 10), which was active for the irregulars and nonce verbs but not for the regulars, leading J et al to speculate that it is responsible for inhibiting the regular, 'default' process. Again, however, the irregular and nonce conditions were harder than the regular condition and harder tasks activate more areas. Moreover, the dual-mechanism theory assumes that irregulars and nonce items are processed by separate mechanisms; any rationale for why they activate the same brain region cannot provide direct evidence for the separate mechanisms, unless it is assumed that separate mechanisms can

occupy the same region of the brain, which is contrary to the logic of using PET to identify distinct functions with distinct loci.

The third area J et al. emphasized is the left dorsolateral prefrontal cortex (area 46), which was active for regular and nonce verbs but not for the irregulars. This finding cannot be due to the relative difficulty of conditions because the regular and nonce conditions were easier than the irregular condition. There is, however, a compelling alternative interpretation of this result for which there is independent neurophysiological evidence. Recall the earlier suggestion that presenting the rule-governed stimuli in homogeneous blocks enabled a simple experiment-specific response strategy because subjects knew in advance that the regular suffix was required. This strategy requires maintaining this information in working memory. Classic studies with monkeys (Goldman-Rakic 1988) implicate the left dorsolateral prefrontal cortex in working memory. The area is active when, for example, a monkey needs to remember a particular spatial location over a delay period. We suggest that the left dorsolateral prefrontal cortex was playing a similar role in J et al.'s study. Neurons in this area of the monkey brain are thought to keep active a trace of the spatial location that the monkey must remember in order to be rewarded. Similarly, in J et al.'s Regular Past and Nonce Past tasks this area of the brain was likely to be involved in remembering that the regular suffix was required.

In summary, the presentation of stimuli in blocks in the J et al. study created differences in degree of difficulty across conditions that contributed to the patterns they obtained. These observations suggest that similar results would obtain for conditions consisting of verbs of a single type (e.g., regular) that exhibited analogous differences in difficulty (because, for example, they varied in frequency).

6. Do dissociations implicate separate subsystems? To this point the question we have addressed is whether J et al.'s data demonstrate differences in how rule-governed verbs and exceptions are processed. We now need to consider whether such differences, should they exist, necessarily favor the dual-mechanism theory. In interpreting their data, J et al. followed the lead

of Pinker, who has said 'Because it categorically distinguishes regular from irregular forms, the rule-association hybrid predicts that the two processes should be dissociable from virtually every point of view.' (Pinker 1991:532) The assumption that only the dual-mechanism theory predicts such differences has led to a vigorous hunt for them in a broad variety of domains. The reasoning underlying the interpretation of such dissociation data needs to be examined closely, however. It is true that the differences between two types of items can be explained in terms of two mechanisms dedicated to processing them, simply by investing each mechanism with whatever characteristics the data demand. However, the converse--that such differences are incompatible with the connectionist approach--is not true. There is an intuition that because connectionist models use a single mechanism to generate all forms, they provide no basis for predicting any differences between the forms that are classified as 'regular' and 'irregular' by the other theory. However, in the networks, verb alternations differ in terms of how hard they are to acquire and process and how vulnerable they are to different types of damage. These aspects of model performance provide a basis for explaining the kinds of dissociations that have been taken as favoring the single-mechanism account. Extensive research on this issue has led to one of the most important findings to have emerged from the first generation of connectionist models of disordered language and cognition: that dissociation data thought to transparently implicate separate mechanisms actually do not (Plaut et al., 1996; Plaut 1997).

The evidence bearing on these issues derives from studies of both the past tense and closely related phenomena concerning the pronunciation of visually-presented words. Studies of verb morphology such as J et al.'s involve presenting a verb and generating its past tense. Studies of word reading involve presenting a word and generating its pronunciation. Both bodies of knowledge are what we have termed 'quasiregular' systems in which most patterns can be described by rules but there are many exceptions that deviate from the rules in varying degrees (Seidenberg & McClelland 1989; Seidenberg 1992). In past tense morphology, there are rule-governed forms such as bake-baked and exceptions such as take-took. In word pronunciation, there are rule-governed forms such gave and mint and exceptions such as have and pint. In both

domains the exceptions tend to be higher frequency items and form overlapping clusters. In both domains generalization is assessed by using nonce forms such as wug.

There are close similarities between the competing theories in the two domains as well. Pinker's dual-mechanism account of the past tense is formally identical to Coltheart et al.'s (1993) 'dual-route' theory of pronunciation. Both consist of a rule component and an associative network used to generate the exceptions. In both domains there is a competing connectionist theory in which all forms are generated by a single network. There is a vast amount of research on normal and disordered word pronunciation and both connectionist and dual-mechanism theories have been worked out in more detail than in the case of the past tense.Given the close similarities between the theories and data in the two domains, the word pronunciation research can provide useful insight about the past tense.

Consider again the J et al. study. Leaving aside the methodological issues raised above, the study yielded different patterns for regular and irregular verbs. Can such data be taken as evidence for two mechanisms? The answer is negative because this 'dissociation' is reliably exhibited by connectionist models that lack separate subsystems dedicated to different types of verbs. J et al.'s stimuli consisted of lists of regular and irregular verbs that were roughly equated in terms of frequency of occurrence in the language. The verbs in both groups ranged in frequency from very common (expected to occur about once every 1,000 words) to much less common (expected to occur about once every 100,000 words). We know from behavioral studies and connectionist models in both the word pronunciation and past tense domains that, other factors being equal, these kinds of stimuli yield an overall advantage for the regulars (Daugherty & Seidenberg 1992; Plaut et al. 1996). In brief, the weights in the network reflect the aggregate effects of exposure to all of the patterns in the corpus (e.g., present-past tense pairs). Performance on any given pattern depends on how it overlaps with other patterns. 'Rule-governed' present-past alternations benefit from the fact that there is a large body of other items that behave similarly ('friends') and few if any that behave dissimilarly ('enemies'). 'Irregular' alternations have fewer friends and many enemies. The weights in the network come to reflect these facts about the ensemble of present-

Running Man 游戏汇总

R unni ng M an游戏汇总 一、铃铛追击游戏 队员分为任务队和追击队。 任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 二、一杯茶的时间 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 三、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 四、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 五、测谎·测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来

六、卡拉OK 每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 七、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 八、1VS n 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 九、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数每个人按顺序说1、 2、 3、4...，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、少站都不行，如果出错就输了 十，乒兵球识字 把字写在乒乓球上，一队打乒乓球，另一队要在这过程中认出球上的

RunningMan策划书

中山大学武警国防生 Running GFS策划书 主办单位：武警驻中山大学选培办 中山大学国防生团委 承办单位：中山大学国防生学生会 二〇一五年三月

一、活动背景 活动构想来源于综艺节目《奔跑吧兄弟》,这是一项丰富课余生活、提升团队合作的活动，希望借此活动加深相互了解程度，让同学们在课业之余放松身心，增强体质。参与者可以通过动脑、动心、团队作战等活动更好地了解团队成员合作的重要性。此项活动形式独特，互动性强，参与度广，既能培养参与者的团队合作精神，又可以丰富了同学们的业余生活。中大国防生学生会积极响应选培办和中山大学国防生团委的号召，计划举办Running GFS活动。 二、活动时间 2015年11月 三、活动地点 中山大学东校区假草训练场 四、主办单位和承办单位 主办单位：武警驻中山大学选培办、中山大学国防生团委 承办单位：中大国防生学生会 五、参赛方式 自愿报名参加，共4支队伍，每支队伍正式参赛队员10人（另有替补队员5人）。 六、比赛内容 【活动一】：首先①②两人（②在下面蒙眼）路过石桥穿过小花园接力给③④，③④两人玩两人三足游戏，转过池塘小石桥去跳绳，至少连续跳10个，最后⑤原地象鼻转10圈，单腿站立5s

结束。以上程序进行两次。用时最短得到线索。 线索一：第一个数字：图片中的战斗机（歼20）属于欧标第几代； 线索二：第一个数字：读一段文字（三国志），该段文字所属书目在 二十四史中的排位。 【活动二】： 根据到操场的速度抽签完成各项活动， 1、在操场找5个同学一起照相，在一张照片里15个人必须同时跳起； 2、在操场找5个同学一起玩跳山羊的游戏，循环一轮； 3、在操场找3个同学，与组内挑选的5个人一起，做到只有2条腿 2只手着地，并停留5秒； 4、在操场找3个同学，与组内挑选的5个人一起，做到 站在对折的报纸上停留5秒。 线索一：第二个数字：是最小的质数； 线索二：第二个数字：(a+a)÷(a*a)=。 【活动三】：假草操场撕名牌大战。在操场各角落总共放置50份资 料，其中有效资料10份。（这些有用的资料里面会有一些问题提示给大家：第三位密码数字是0，第四位密码数字是1.）在这个过程中，不同队伍之间可以相互撕对方名牌，被撕者退出游戏。 【最后】到网球场打开宝箱，密码4201。 七、比赛安排 （一）比赛共设4名裁判，4名记分员，4名计时员 a 0a 1dx

2021年Running man 游戏汇总之令狐采学创编

Running man 游戏汇总 一、 欧阳光明（2021.03.07） 二、铃铛追击游戏 队员分为任务队和追击队。 任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 三、一杯茶的时间： 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 四、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。五、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 六、测谎·测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定

一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 七、卡拉OK 每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 八、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 九、 1 VS n 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 十、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数每个人按顺序说1、 2、 3、4...，那就不能同时站起来两 个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、

三走趣味活动策划书

三走趣味活动策划书 一、活动名称“走下网络、走出宿舍、走向操场、走进图书馆”之“科普小站runningman ” 二、活动单位 主办单位：共青团xx 学院委员会xx 学院体育学院承办单位：xx 建设大学生科技服务团协办单位：红十字会秘书长助理团 三、活动对象 xx 学院全体学生 四、活动时间 初赛：XX年4月18日 决赛：XX年4月19日下午14:00 注：如遇下雨或其他安排则顺推一天，以短信通知为准。 五、活动地点 初赛：xx 学院奇迈篮球场、二三组之间大道、四组食堂空地、东区田径场 决赛：xx 学院、同心广场 六、活动内容 一初赛 a）时间：XX年4月18日

b）地点：xx 学院奇迈篮球场（签到） c）活动流程 1.团体接力大： 地点：xx 学院奇迈篮球场 规则： 队伍中4人，a进行投篮直到投进3球后与b击掌；b进行竞走（过程中只能竞走不可跑，否则视为犯规），竞走到终点后与 c 击掌；c 开始跳绳10 个与 d 击掌， d 将饼干放于额头，不能用手而吃到饼干。全程计时。 由工作人员统计算出用时最短的前六个队伍。本轮游戏获胜者可抽取对接下来游戏有利的条件，第一名的队伍可抽 3 次，二三名抽 2 次，四五六抽 1 次。（有利条件包括：pass 卡、求助卡、 提示卡） 2.我和路人有个约定： 地点：二三组之间大道 规则： 参赛队伍寻找路人并带到工作处，领取 2 枚色子。每名队员都与路人进行掷色子，四人必须有 2 人赢方可进行下一轮游戏，否则要马上寻找另一个路人，重复以上规则。完成这一轮的任务后， 工作人员粘贴特殊符号粘纸。

3.猜、猜、猜：地点：四组食堂背后的空地（四组 1 栋和 2 栋 之间的空地）规则： 参赛队伍，每人抽取一张纸条（纸条里有 3 个或 2 个题目或 1 个题目）并回答问题，全猜中后，工作人员粘贴特殊符号粘纸。 4.决赛队伍评选：地点：东肖操场的榕树下工作人员粘贴决赛参赛资格贴纸，直到8 张本轮参赛资格纸贴完为止，即先到的前8 个队伍方可进行下一次的决赛，其他的队伍淘汰。 二决赛 a）时间：XX年4月19日下午14:00 b）地点：东区田径场大榕树下（签到） c）活动流程 1. 解密： 地点：xx 学院

Runningman游戏大全

Running man 游戏大全 1、泥潭 抢裤子，穿上跑向终点 在泥池里推人出来，留在里面的人win。 泥潭赛跑or接力 2、撕名牌 铃铛追击 攻受交替 反射复活各种 1对3 每组成员选出自己的组王，与其他组的另3名成员去撕名牌，如被撕掉，比时间长短，如撕掉，则被撕掉的组直接out。 不知情下撕名牌。 撕名牌的同时完成任务 3、拍照 10秒钟照相，自己按快门，做规定动作。 环境可变，要求各种动作等等，最终出现在镜头里的队获胜。 4、指压板 跳绳or接力or比跳远的远度距离or互追各种完成任务的比赛 5、一杯茶时间tea time 先通过骰子或勺子确定饮品的量，比如鸡蛋，柠檬，咖啡等等，再选择加入调料（如糖等等）的勺子（汤勺，米勺，调羹，小勺子，挖耳勺等等），最后比赛接力喝完。先喝完win。 6、接力 跑步接力加各种障碍or任务 各种各种自己也可以think and create 7、趣味KTV 你懂得哈哈哈各种搞笑 8、一对多 在一个设计好的游戏中，众人欺骗一个人，到结束时不被揭露则win。 9、水池 飞椅系列（读绕口令答题完成任务各种） 跳水比高度总和or比远度总和。 人体保龄球 垫上推人 草垫冲刺加垫上推人 10、躲避球 情侣躲避球 保护女王 三队之间的躲避球 11、绳（可加上各种小任物，好玩，又有挑战） 可以加上找人（有限制如：国籍身高性别学校姓氏各种）跳大绳集体N人，跳N个win

双人跳绳跳多少个win，或集体加到一块跳N个win。 三人跳绳同双人，or 中间一人顶着装有水的杯子，跳N个，然后水要剩在线之上，则win。 大绳多角拔河三组or多组，敲锣or抢东西。 12、无穷花开 选一个人当鬼在鬼一句话说完后就回头大家就都不能动谁动了就会被抓然后和做鬼的人领着手直到有人到鬼的身后并在鬼没发现的情况下 打开他们领着的手然后鬼追大家跑被追上的就当下一个鬼如此循环 如果是“情侣”战的话，也就是背人，可以让鬼单脚跳追人。 各种情景设置 Running man 基本每一期都会有一个故事背景，感觉特别好！ 各个历史各个国家电影故事各种very good！ 如双重间谍、水枪、寻宝、前生今世、 澳门站：找箱子招钥匙挑战（找三个字母，组成一个词）蹦极223米，百步登天，空中漫步、找人（有限制）跳绳。 平时易实施的 1、两人扔一个球，转身后接着为win。 2、吃饼干or其他，在脸上或者杆上或者帽子上吊着。 3、衣服夹子夹脸拔河。 4、用手折断筷子。可以把筷子夹在中指上，食指和无名指下，拍桌子折断即可。 5、找人，让人和自己一起戴帽子（用线连着），然后对视10秒。 6、传递接力（薯条和圈） 7、碰鸡蛋幸运1个生鸡蛋和N个熟鸡蛋，挑鸡蛋用头打破，打到生鸡蛋的out。 8、竖鸡蛋竖易拉罐瓶 9、表演猜可变通，各种领域和方法，成语，电影，音乐，明星等等。（这个好玩）（表演的时候可以有限制，比如在蹦床上，双杠上等等） 10、将纸巾吹到篮子里。 11、喝完奶后把瓶子扔到垃圾桶里。 12、抢东西然后传递。 13、穿上脚蹼，三人接力绕过障碍，多少秒之内完成win。 金字塔前进（成跪姿） 14、看眼色游戏 一人报数，其余人按所报数字站起相应的人数，出错的out。 先蹲下，然后每个人按顺序报数并站起，最后一个报数或者同时报一个数的out。 15、 16、记歌词 播一首歌，小组能完整，准确唱出的win。 17、翻勺子 上面写上金银或者其他，或者勺子不同。然后比谁先翻错。

runningman游戏汇总

Running man 游戏汇总 一、铃铛追击游戏 二、队员分为任务队和追击队。 三、任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 四、追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 五、一杯茶的时间： 六、两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用 该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 七、照相 八、要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照 相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 九、记歌词 十、播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 十一、测谎·测心率 十二、一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个

谎言，通过测谎仪来评定 十三、一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 十四、卡拉OK 十五、每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK 的评分来定输赢 十六、画画传递猜词? 十七、一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 十八、 1 VS n? 十九、众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 二十、眼神游戏 二十一、大家都蹲下来，然后一个一个地站起来。 二十二、1、报数每个人按顺序说1、2、3、4...，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起

runningman游戏汇总

一、铃铛追击游戏 队员分为任务队和追击队。 任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 二、一杯茶的时间： 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 三、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 四、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 五、测谎·测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 六、卡拉OK 每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 七、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜

八、 1 VS n 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 九、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数每个人按顺序说1、 2、 3、4...，那就不能同时站起来两个人，如果站起来两 个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、少站都不行，如果出 错就输了. 十、乒兵球识字 把字写在乒乓球上，一队打乒乓球，另一队要在这过程中认出球上的字 十一、倒水游戏 队中选出一人蒙眼往杯中倒水，其余队员坐在椅子上，将杯子举在头顶上，最后水多的队获胜 十二、衣架接力赛 十三、369BIBO 几个人按顺序数数，但当遇到含有3,6,9的数要用BIBO代替，否则输掉 十四、模拟恶劣天气限时播报 三个人均给出播报词，但在另一地点只提供三套衣服，队员要在抢衣穿衣过程中记住词，并在人为制造的如，大风，大雨等恶劣情况下正确播报。 十五、来了两个人 几个人按顺序蹲成一排，另一个人随意说数字，那几个人要按照顺序站起相应人数 如甲乙丙丁四人， 喊“1”,则甲站起来；接着喊2，则乙和丙站起；再喊2则丁和甲站起 十六、集体跳大绳敲锣

running man策划书

南京工业大学 Running man 策 划 书 主办单位;共青团南京工业大学委员会 承办单位;共青团南京工业大学能源学院委员会活动时间;二0一六年十月十六日

能源学院Running Man策划书 活动背景 新型Running Man运动是一项挑战智慧与体能的运动，考验人们在不同环境压力下能否做出对路线的正确判断和是否具有果断的决策及应变能力。参与者借助地图，以徒步越野跑的形式，在最短的时间内跑完每个检查点并通过关卡设置后后到达终点。在每一个地点过后会告诉你下一个地点。跑步会让你每天都自我感觉良好：我自己的感受是，即便我自己不想跑的日子，在费了好大的勇气，让自己穿上跑鞋，走出家门，一路跑下来后，我总会感觉良好，做事更有效率。所以一起来加入我们吧！ 一、活动目的 增强同学们的体质，培养同学们的运动意识和协作精神，丰富校园生活。 二、活动主题 快乐运动健康生活 三、组织机构 主办单位：共青团南京工业大学委员会 承办单位：共青团南京工业大学能源学院委员会 四、活动时间 二0一六年十月十六日 五、活动流程

1、活动宣传 能源学院大学生科协对外宣传部送发活动策划书至各学院学办并在东苑食堂和亚青村进行设点宣传。 2、活动报名 本活动将以组队参赛形式展开，每队三人,请各位参赛选手自行组队完毕后填写活动报名表并及时与对外宣传部相关人员联系，联系方式附在报名表上。 报名截止日期：2016年10月9日 3、活动展开 本次活动我们将以东苑篮球场为中心，在周围设置六个地点，每支参赛队伍在东苑集合，待所有队伍到齐后开始比赛。比赛开始后，工作人员将以某种特殊方式告诉每支参赛队伍需要到达的地点，然后队伍开始run，到达指定地点后，我们工作人员会设置有闯关或者答题环节，例如你画我猜，我爱记歌词，N人N+1足，定点投篮等等，参赛选手要以最短的时间完成题目或者完成闯关项目（期间讲解闯关注意事项时间不计入总时间），完成后工作人员将会告知下一个地点，并且会给选手一份拼图的一部分，待六个地点各个队伍都完成后再回到东苑篮球场，将拼图组装完成后完成比赛，此时计时完成。 4、活动评比 从出发点出发开始计时，到拼图完成计时结束。根据比赛时间以及各个队伍完成闯关情况分出名次。

奔跑吧兄弟户外小游戏活动方案

奔跑吧兄弟户外小游戏活动方案 一、活动前言 韩国超火综艺节目“running man”由于它强大的娱乐性和竞技性，吸引了全世界的注意力，可以说是风靡全球，然而坐在电脑前的我们无法身临其境，只能享受视觉效果，为之失落不已。然而，今天，地理学院即将为大家提供尽情表现自我的机会，“Running star”专属于江西师大的校园版“Running man”，重磅来袭!这里没有璀璨夺目的明星，但是有青春洋溢的我们;这里没有华丽绚烂的场景，但是有绿意盎然的健康小道;这里没有专业大牌的导演，但是有认真服务的小编。一段段精彩的演出将会由你来呈现，这将会是你的专属show time! 二、活动目的 为了提高大学生的综合素质，发扬团队协作互帮互助的精神，使广大学生在活动参与中受到潜移默化的影响，思想感情得到熏陶，精神生活得到充实，道德境界得到升华，活跃我校大学生文化体育生活，营造良好的校园文化氛围，加强各分站间的交流。 三、活动主题 Running star 四、活动主办方 地理与环境学院

五、活动承办方 软件学院、传播学院、生命科学学院、体育学院、理电学院 六、活动时间 十一月八日下午两点 七、活动地点 健康小道 八、主持人 地理与环境学院文化部部长 九、参与人员 (一)嘉宾 地理与环境学院党员服务站站长 软件学院党员服务站站长 体育学院党员服务站站长 传播学院党员服务站站长 生命科学学院党员服务站站长 理电学院党员服务站站长 (二)参与人员主体 地理与环境学院党员服务站内部人员 软件学院党员服务站内部人员 体育学院党员服务站内部人员 生命科学学院党员服务站内部人员

传播学院党员服务站内部人员 理电学院党员服务站内部人员 十、活动安排策划书 (一)活动前期工作 1.由地理与环境学院党员服务站文化部负责写活动策划 2.由主办方和承办方各派出5名工作人员 3.由各学院党站办公室人员负责统计报名人数，最后地理学院服务队办公室总计 4.由各学院服务队实践部负责准备活动所需道具 5.由地理学院党站宣传部负责设计任务完成的标志，并负责拍照宣传 6.参与的各学院邀请参加此次活动的嘉宾 7.由各学院文化部部长向站内人员宣传，并讲明游戏规则，由地理学院服务队办公室随机组队，各组人员赛前互相熟悉，并想好队名、口号。 (二)活动当天流程 1.工作人员提前半小时去健康小道布置任务点，参与人员提前15分钟到场 2.工作人员组织随机分组，共16个任务点，两人一组一个任务点，起点和终点各有一名计时员。 3.主持人讲话，包括对嘉宾的介绍，主办方、承办方的

runningman成功因素分析

黄佳佳 131504012 13广告韩版runningman的独特魅力及我喜欢它的原因《Running Man》是韩国SBS电视台周末娱乐节目，节目致力于打造一个不同于过去real variety的新型态娱乐节目。每期由7位固定成员及不同嘉宾参演。对应每期节目不同的主题，分为不同的队伍进行比赛，最后获胜一方将获得称号或奖品。游戏中创造的撕名牌等环节收获无数好评，也使得许多综艺节目竞相模仿，其中在中国影响最深的就是邓超、Angelababy、郑恺、李晨、王祖蓝、陈赫以及包贝尔出演的《奔跑吧，兄弟》都已更新至第二季，其独特的魅力使得它在全亚洲都有极高的人气。《running man》是韩国sbs电视台于2010年推出的全新游戏类综艺节目，由刘在石、金钟国、哈哈、池石镇、李光洙、GARY、宋智孝等7位高知名度的艺人主持。节目定位幽默风趣，积极向上。刚开播的时候，游戏类型大多比较简单，其中防守游戏是节目的核心，嘉宾阵容以演艺界和歌谣界为主，选取的场所多数是国内一些标志性的建筑物和旅游景点。自11年开始，场所开始逐渐向海外拓展，泰国之旅是开启RM海外市场的第一步，随后北京、香港、澳门逐一成为节目的取景点。2012年开始，游戏形式开始实行突破，以完整故事背景的游戏模式让人眼前一亮，更多电影元素，文化元素开始融入游戏中，嘉宾的选择也不再仅限于演艺界，2012之后的RM，无论在内容上、形式上还是嘉宾选择上都趋于多样化，源源不断的探索和创新也成就了如今RM有着至高的人气。 总结其独特而成功的魅力原因有以下几点： 一、巧妙的人员组合 RM共有七位固定主持（数据以现在为准，中间数次人员微调动除外），一女六男，分别来自歌谣界、演艺界和综艺界，以韩国号称综艺两座大山的刘在石为头形成了一支多才多艺，各具特色的主持队伍。不难看出，在相貌的筛选上，男主持人大部分只是中等甚至中等偏下，其中几位在街上完全是路人甲乙，在视觉上并非盛宴。因此常常可以听见主持人相互之间也会拿平庸的长相说笑甚至自嘲，在这过程中主持人在言语间和语调上处理得恰大好处不但不会引起道德层面的冲突，反而给节目的增添了更多幽默色彩。唯一的女主持宋智孝的确是美女，这很好地抓住了男性心理特征，而宋智孝的定位本身也不是搞笑，从现在的头衔“ACE懵智”就能看出，宋智孝所要塑造的是一位美貌与

runningman超好看的期数

100912 E10 嘉宾：车太贤、尹世雅任务：寻找running ball 地点：国立现代美术馆 蓝队受罚（除宋智孝，在RM拍卖中获得免除惩罚权） 惩罚内容：在带有惩罚妆底下在弘大便利商店玻璃前吃泡面 100919 E11(缺宋仲基) 嘉宾：郑容和（C.N.Blue）、金济东任务：获得running ball（个人战） 游戏：骗Gary 8vs1(Gary)、明信片抽小偷、选择包裹箱子、字尾接龙、捉迷藏（找密码） 刘在石、李光洙、金钟国、Gary受罚 100926 E12 嘉宾：无任务：获得running ball（个人战） 游戏：设计竞猜7vs1（宋仲基）、明信片抽小偷、设计阶梯、眼神游戏、捉迷藏（解除炸弹） 池石镇、刘在石、李光洙、宋仲基受罚 101031 E16 嘉宾：Yuri（少女时代） 任务：获得running ball（个人战） 游戏：计时猜谜8vs1（Yuri）（running man 胜）、试吃三人组队体重增加1Kg、捉迷藏（收集20万发票） Yuri、李光洙、金钟国、Gary受罚 101107 E17 嘉宾：郑容和(C.N.Blue)、高周元任务：获得running ball（个人战） 游戏：听写1（郑容和）vs9、组Band 外援音乐专业女大学生选择【全部选错】、演奏猜曲、捉迷藏（设计官找摇铃） 宋仲基、刘在石、池石镇、Gary、高周元受罚 嘉宾：Nichkhun（2PM） 地点：南山韩国之家 任务：获得running ball（个人战） 黄队：李光洙、宋智孝、宋仲基蓝队：Nichkhun、HaHa、金钟国、Gary 红队：刘在石、池石镇、Lizzy 处罚：宋钟基、刘在石、池石镇、宋智孝、李光洙、Lizzy 101212 E21 嘉宾：金济东 任务队：刘在石宋钟基池石镇 Lizzy 哈哈Gary 李光洙金济东 捕捉队：金钟国宋智孝 地点：高铁幸信站、龙山站、光明站 任务：获得running ball（个人战） 游戏：搭乘列车、1（金济东）vs9【金济东胜】、捉迷藏

running man操作流程

操作流程： 1、按报名先后顺序安排分组。 2、布置任务：终极任务是找到一个“宝地”，并且全队人员用手机和宝地有一 个合影。要先通过无数的关卡，完成关卡任务才能得到一个提示，根据提示找到下一个关卡，最终找到终极“宝地”的提示。最先找到的队伍能获得一份神秘大奖。 3、出发地：篮球场（第一个任务地） 第一个任务：30秒内每组累计跳绳120个。其中一个队员“甩贴纸”，甩完才能得到提示卡。 提示卡：到达餐厅门口 负责人： 4、餐厅门口（第二个任务地） 第二个任务：两个人30秒内背人快速通过15米，再背回来。两个人各连续7次立定跳远，总距离超过30米。 提示卡：到达车间门口 负责人： 5、车间门口（第三个任务地） 第三个任务：抽签俯卧撑（抽到几个点，全队就用几个点支地做10个俯卧撑）（8手0脚或8手1脚或8手2脚） 跳关追逐赛 提示卡：到达活动室门口 负责人：

6、活动室门口（第四个任务地） 第四个任务：每个人依次顶气球，并扎破。托10个球乒乓球行进，5秒内跑完15米，剩下的球不能少于5个。 提示卡：到达西餐厅门口 负责人： 7、西餐厅门口（第五个任务地） 第五个任务：飞镖必须1人1标，总环数超过N*8+1环。三次机会后需派人做10个俯卧撑才能获得第四次机会……若第二组到来后，各组需轮换进行。 提示卡：到达南门口 负责人： 8、南门口（第六个任务地） 第六个任务：两个人打保龄球；两个人把饼干从眼睛上通过脸上肌肉运动到嘴里。 提示卡：到达1号门口 负责人： 9、1号门口（第七个任务地） 第七个任务：吞水传音 提示卡：到达停车场 负责人： 10、停车场（第八个任务地） 第八个任务：翻页子 提示卡：到达停车场

Runningman游戏汇总

R u n n i n g m a n游戏汇总1、铃铛追击游戏 队员分为任务队和追击队。 任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 2、一杯茶的时间： 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 3、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 4、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 5、测谎·测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 6、卡拉OK 每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 7、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 8、1VSn 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 9、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数每个人按顺序说1、 2、 3、4...，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、少站都不行，如果出错就输了.

分析比较Running man 和跑男

分析比较Running man 和跑男 一、总体介绍 1、基本情况： RM：2010.7.11开播，截止至2015.3.2，共计播出236期。韩国SBS电视台《星期天真好》单元的综艺节目。Running man 这个节目已经做了好多季了，有几百期了，在韩国的综艺界有着至关重要的地位，一直特别深受广大韩国观众的喜爱，同时也成为韩国其他综艺节目借鉴的范例。 跑男：2014.10.10开播，截止至2015.1.23，共计播出16期。中国浙江卫视大型户外竞技真人秀节目。 2、获奖情况： RM：2010-2014，连续五届SBS演艺大赏获奖节目 跑男：16期节目中14次获得同时段收视率第一 3、制作单位： RM：SBS株式会社，1990年成立，韩国三大电视台之一，韩国国内唯一覆盖全国网络民营无线电视台。出产大量娱乐综艺节目以及电视剧，受年轻人欢迎，先后制作过30余个娱乐节目，例如强心脏、X-man、人气歌谣、家族诞生。 跑男：浙江卫视，1960年成立，省级卫视前三强，高影响力，全省覆盖率97%。制作过十几个节目，其中中国好声音、十二道锋味、我爱记歌词为大家所熟知。 4、两者之间的关联性： ①跑男是由韩国的running man 引进来，无论是主持人的定位设置，还是游戏环节的设置，都保留了很多原版的特点，如趾压板和撕名牌。有几期节目甚至把原版的叙事结构直接复制过来，如邓超成为内奸那一期。 ②两个节目的结构也大体相似：每期都会有一个固定的贯穿始终的主体线 如：跑男：寻找白蛇、高校争霸战等 Running man :公主和他的骑士们、定制情侣等 围绕这个主体线，以此为依据分组（跑男一般是分为两队，有时候三队，极少数时两人一组等；running man 则基本都是三人组来分队），然后各组像闯关一样开始比赛，每关赢得比赛的组可以获得一个提示或者特权等，一直进行到最后，开始撕名牌环节的比拼。 5、跑男影响力 当日发布的全国34城周五档上星频道重点节目收视率显示，《奔跑吧兄弟》成为出央视外，其他省级卫视收视的总和。 值得一提的是，“跑男”十二期节目平均收视率2.41，超过江苏卫视《非诚勿扰》的平均的收视率2.389与其他三档节目的领先优势也在不断扩大。反观开局良好的《明星到我家》收视率起伏颇多，且呈现出下滑趋势，《一年级》和

首届“奔跑吧”runningman挑战赛活动方案

首届“奔跑吧”runningman挑战赛 活动背景： 浙江卫视的综艺节目“奔跑吧，兄弟”由于它强大的娱乐性和竞技性，可以说是风靡全国，其中的“撕名牌”环节更是让许多观众跃跃欲试，首届“奔跑吧”runningman挑战赛即引入此项互动性强，参与度广的活动，既能培养员工的团队合作精神，同时又丰富了员工的娱乐生活，更是发扬全民参与、全民健身精神的大好时机。 活动目的： 趣味挑战赛是娱乐与健康相结合的游戏。平常员工都忙于工作而忽略了身体的锻炼，趣味挑战赛给他们提供了锻炼的机会。在比赛中也能够感受到运动的魅力以及运动的重要性。通过各种趣味挑战的环节，令员工忘却日常工作的烦恼，全身心投入娱乐中来，为重新工作积蓄活力，同时也会培养员工之间团结协作、良性沟通等素质，对企业产生强烈的好感和信任感，激发团队更高昂的工作热情、拼搏创新精神，让团队更富有凝聚力，深化企业文化。 活动主题：XX集团，活力飞扬！ 活动时间： 活动对象：公司各管理人员及后勤人员。 活动地点： 活动筹备单位： 活动流程：

（一）前期准备： 1、报名时间：2014年5 月30 日18：00之前 2、报名方式： 参赛者将报名信息（姓名、联系方式、部门）发送给活动报名负责人进行报名，进行人数统计。 3、分组抽签： 活动为团体赛，参赛人员分为筹备组、裁判组、红队、蓝队。红蓝队按照年龄层次以及男女比例进行分组抽签，红队14人，蓝队14人。 （二）活动内容： 赛制说明：比赛分为四关进行，前三关每一关获胜的队伍由裁判组成员颁发一颗万众球，第三关结束后，集齐万众球最多的队伍获得第四关撕名牌环节的主动权，掌握主动权的队伍则是第四关的追击队。 第一关：智慧大比拼 你划我猜： 红蓝两队各将14人分成两组，两人用肢体语言比划词语或成语，一人猜，在限定时间内答出最多词语的一组即算获胜，获胜队将由裁判组成员颁发一颗万众球。 第二关：默契大考验 折返跑： 红蓝队两人分组，50米折返跑。两人三足起跑，到达折返点后

RunningMan赞助策划书

哈尔滨工业大学 隐秘而伟大 哈工大第二届RunningMan 综合挑战赛 赞助策划书

目录 .主题背景 (3) .活动概况 (3) 基本信息 (3) 活动要求 (3) 活动流程 (3) .宣传计划 (4) .市场分析 (4) .赞助补偿计划 (5) 冠名权 (5) 平面宣传 (5) 网络宣传 (5) 广播媒体宣传 (5) 视频媒体宣传 (5) 现场宣传 (5) 鸣谢 (5) .工作明细 (5) .应急预案 (7) .经费明细 (8)

主题背景 活动主题：哈工大第二届RunningMan综合挑战赛 承办单位：哈尔滨工业大学学生会 活动背景： RunningMan原本是一档韩国的综艺节目，以做任务、找线索、解谜为主的活动。根据调查，在工大有很多收看RunningMan这个节目的学生。市政环境工程学院文艺部已成功举办第一届RunningMan，为了让更多的学生切实体会到这种活动的魅力，本部决定举办第二届RunningMan综合挑战赛。 活动目的： 在RunningMan这个活动中需要体力和脑力的配合，可团体合作亦可单枪匹马挑战任务。本期RunningMan设计四个游戏环节，考验参赛者的应变能力，并使其体验团队精神的力量。 活动概况 1.活动时间：四月上旬某周六上午

2.活动地点：阳光大厅、排球、教学主楼 3.活动流程：前期宣传（海报、视频、媒体等）→现场宣传选拔 →活动举办→后期总结宣传 宣传计划 一、前期宣传 1、网络宣传：人人，新浪微博，微信等网络媒体的宣传 2、校园广播宣传：与校园广播合作，定期进行品牌宣传 3、平面宣传：传单，海报，条幅等各种形式的宣传 4、视频宣传：在食堂等人数集中的公共场所进行视频宣传 二、活动宣传 1、现场宣传：场宣的招募宣传 2、活动宣传：挑战赛进行中的形象宣传 市场分析 —本次活动应用名牌效应，吸引人数庞大。同时，哈工大为知名重点大学，有着足够多的关注群体，经过这一活动的宣传，必定能有效提升赞助商知名度； —与电视、报刊等相比，学校宣传有极高性价比，可用最少资金做到最好的宣传；

Runningman游戏汇总

Running man游戏汇总 -、铃铛追击游戏队员分为任务队和追击队。任务队要完成一定的任务，同时保护自己的名牌不被撕掉；追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 二、一杯茶的时间 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢（越夸张越好）然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉）然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 三、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前， 有时照相机位置可移动，且照相环境可变（如在水中、舞台上）当照相机照下照片后，露脸多的队获胜。 四、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜五、测谎?测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 六、卡拉0K 每队在完成各种任务（如坐过山车、举重、压腿）的情况下唱歌，通过卡拉0K 的评分来定输赢 七、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜八、1 VSn 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装又不能让其发现。若被揭露, 则被欺骗者获胜。 九、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数：每个人按顺序说1、 2、3……，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3” ,就要有3个人站起来，多站、少站都不行，如果出错就输了 十、乒兵球识字 把字写在乒乓球上，一队打乒乓球，另一队要在这过程中认出球上的字

最新Running man 游戏汇总

Running man 游戏汇总 1、铃铛追击游戏 队员分为任务队和追击队。 任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 2、一杯茶的时间： 两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 3、照相 要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 4、记歌词 播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 5、测谎·测心率 一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 6、卡拉OK 每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 7、画画传递猜词 一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 8、1 VS n 众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 9、眼神游戏 大家都蹲下来，然后一个一个地站起来。 1、报数每个人按顺序说1、 2、 3、4...，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、少站都不行，如果出错就输了. 10、乒兵球识字

Runningman游戏汇总

R u n n i n g m a n游戏汇总 一、铃铛追击游戏 二、队员分为任务队和追击队。 三、任务队要完成一定的任务，同时保护自己的名牌不被撕掉； 四、追击队要在鞋上系着铃铛，同时追击任务队，撕掉任务队的名牌。 五、一杯茶的时间： 六、两队通过抽签选择出不同的器皿，如耳屎勺，汤勺，瓢...（越夸张越好），然后用该器皿向杯子里 装一勺的材料（如咖啡粉、茶粉...），然后泡开，有时可再加入如一颗鸡蛋等作料，比哪队先喝完。 七、照相 八、要求各队所有人想尽办法让自己的脸呆在不知何时会拍照的照相机镜头前，有时照相机位置可移 动，且照相环境可变（如在水中、舞台上），当照相机照下照片后，露脸多的队获胜。 九、记歌词 十、播放一首歌曲，各队要仔细听，能完整、正确地唱出该歌曲的对获胜。 十一、测谎·测心率 十二、一队中选一个人连接上测谎仪，另一队人要想办法让那一个人在规定时间内说3个谎言，通过测谎仪来评定 十三、一队中选一个人连接上心率仪，另一队人使用各种手段（或揭短，激怒，使兴奋）让那一个人在规定时间内心率达到最高，通过测谎仪来评定 十四、卡拉OK 十五、每队在完成各种任务（如坐过山车、举重、压腿...）的情况下唱歌，通过卡拉OK的评分来定输赢 十六、画画传递猜词 十七、一队人按顺序排成一列，第一个人依据所给出的主题作画，并传给第二个人（不能有语言交流），再由第二个人根据看到的画重新作画，并传给第三个人，以此类推，若最后一个人能猜对主题，则该队获胜 十八、 1 VS n 十九、众人合伙要在一个实现安排好的假游戏中欺骗某一个人，既要达到某种目的，如不能让被欺骗者在游戏中获胜、众人集体秘密变装.......又不能让其发现。若被揭露，则被欺骗者获胜。 二十、眼神游戏 二十一、大家都蹲下来，然后一个一个地站起来。 二十二、1、报数每个人按顺序说1、2、3、4...，那就不能同时站起来两个人，如果站起来两个人那就是这两个人输了，要互相观察彼此，看到没人站起来或是马上站起来抢先，因为没有提前串通，所以就很小心的看别人眼色行事，剩下最后一个没站起来的人就输了 二十三、2、游戏外的人喊数，比如喊“3”，就要有3个人站起来，多站、少站都不行，如果出错就输了. 二十四、乒兵球识字 二十五、把字写在乒乓球上，一队打乒乓球，另一队要在这过程中认出球上的字 二十六、倒水游戏 二十七、队中选出一人蒙眼往杯中倒水，其余队员坐在椅子上，将杯子举在头顶上，最后水多的队获胜 二十八、衣架接力赛 二十九、369BIBO 三十、几个人按顺序数数，但当遇到含有3,6,9的数要用BIBO代替，否则输掉 三十一、模拟恶劣天气限时播报