Here is what Keith says
The late Robert Thouless, former President of the Society for Psychical Research, developed a test of survival where a message is encrypted in such a way that it can only be decoded by key words known only to one who has died (Stevenson 114). Thouless set up three encrypted messages for himself, hoping to communicate the key words which would decode his messages to his colleagues after his death through a medium. Though the first cipher he proposed was cracked a few weeks after he published it, neither of the other two ciphers were broken during his lifetime, providing a rare opportunity for parapsychologists to produce truly compelling evidence for survival of bodily death. The key to one of the remaining ciphers (a replacement for the cracked key) was a simple two-word key; the key to the other was a roughly 100-word literary passage. The literary passage key, though long, could be obtained simply by relaying the title of the book, the location of the passage in that book, and a couple of words from the beginning of the passage (Oram 118).
Under the advice of Ian Stevenson, Thouless also transposed the first six letters of his two-word key into numbers using a published table in order to reset a combination lock to those numbers (Stevenson 114). Unlike Thouless' encrypted message tests, the combination lock test requires the entire key to be known to unlock it and gives no hints that one is coming close to hitting the key by making near-misses, thus ruling out the possibility that one could narrow down his options for a key by repeated attempts to unlock the combination lock (115). Stevenson has reported that the odds of hitting the right key for a combination lock test purely by chance are 1 in 125,000 (115).
When Thouless died in 1984 roughly one hundred candidates for a key were submitted to the Society for Psychical Research, some of which came from mediums, but none of them were able to decipher any of Thouless' encrypted messages (Stevenson 114). However, in 1995 James Gillogly successfully decoded one of Thouless' messages using the two key words "black beauty" generated by a computer program he wrote which yielded the message: "This is a cipher which will not be read unless I give the key words". The discovery of Thouless' two-word key was further confirmed when Stevenson used his table to transpose "BLACKB" back into numbers which unlocked Thouless' combination lock (115).
The flaw in Thouless' test which allowed it to be deciphered was his use of common words for a key that could easily be cross-checked by a computer program designed to form two-word combinations from all the entries in a typical dictionary (Oram 116). Incredibly, the parapsychologist Arthur Oram has come to the credulous conclusion that the repeated failures of mediums to come up with a key that will successfully unlock Thouless' encoded messages despite numerous tests is due to the inability of the deceased Thouless to remember the simple keys 'on the other side'! (Although the deceased Thouless could apparently remember who Oram was and other similar facts) (Oram 117). A simpler explanation for these failures is that Thouless could not communicate the key words because he had not in fact 'survived' his death and thus was not in contact with Oram through the mediums. To his credit, Oram does concede this point:
It seems fair to assume that if [the mediums] were in effective touch with Thouless they would either be given the key or an explanation that [and why] he cannot remember or cannot communicate the key ... [A] considerable number of people have felt that they have been in touch with Thouless and some of them have felt that rather deeply, including at least one whose submitted key was of a wrong form [italics mine] (117).
Regarding Thouless' simplest test Oram reports: "There are no instances in our records of anyone getting the two-word key even partially correct" (118). Other similar direct tests of the survival hypothesis have also yielded negative results: attempts to obtain postmortem Thouless' literary passage key to his remaining message, J. Gaither Pratt's mnemonic key to his combination lock, and T. E. Wood's key to his enciphered message have all been unsuccessful (Stevenson, et al, 329-334). Oram succinctly characterizes the state of the experimental evidence for survival from mediumship: "We can only be sure of two facts relating this research; one is that work has been done to try to get the keys through mediums and the other fact is that we have not obtained the keys" (Oram 118).
My response: Apparently, deceased Thouless could remember Oram and other similar facts. Now Keith says even though that is so the simpler explanation would be because he didn't remember the simple keys would be that he didn't survive his death. But their are certainly more explanations then that. One explanation would be that he was not communicating with thouless at all but some other entity and was being deceived by this entity. To me the explanation that another entity who pretended to know Oram is more plausible than the explanation that Thousless didn't survive death. Why? because apparently this entity knew Oram, did Oram ask this entity if he could know anything that thousless would know when he was alive?. According to Oram this entity just knew him. So what we have hear is no confirmation that this was indeed Thousless or some other entity pretending to be Thouless.
Of course Keith doesn't mention in his essay the case against immortality, strong evidence that apparently point to survival. Such as the cross correspondences, another is the r-101 case, the book tests, newspaper tests, along with Ian Stevenson's 21 best cases for reincarnation etc.
The r 101 case for example can be read here