The Algorithm Didn’t Like My Essay

It does not necessarily lead to an arms race. Once the students learn to write to the gradind software, the schools can claim the students are getting better, and point to an 'objective, unbiased' measuring stick. Everyone wins except for society, i.e. sll of us.

It should be built-into word processors, so all students start with A grammar.

The biggest problem I see is, in the real world, concise is king... and marketing is the epitome of persuasiveness.

Microsoft marketing is God-awful fluff, but I guess that's where most consumers are -- stupid.

God says... 25:47 And if a sojourner or stranger wax rich by thee, and thy brother that dwelleth by him wax poor, and sell himself unto the stranger or sojourner by thee, or to the stock of the stranger's family: 25:48 After that he is sold he may be redeemed again; one of his brethren may redeem him: 25:49 Either his uncle, or his uncle's son, may redeem him, or any that is nigh of kin unto him of his family may redeem him; or if he be able, he may redeem himself.

25:50 And he shall reckon with him that bought him from the year that he was sold to him unto the year of jubile: and the price of his sale shall be according unto the number of years, according to the time of an hired servant shall it be with him.

25:51 If there be yet many years behind, according unto them he shall give again the price of his redemption out of the money that he was bought for.

----

My India-nigger doctor probably thinks a college English professor knows more about persuasive writing than a Coke marketing manager.

God says... stoked Mozambique thank_you_very_much you're_no_fun impossible wastoid ghastly fer_sure Honduras let_me_count_the_ways I'm_off_today obviously special_case Israel what_a_mess resume freak fool Malaysia by_the_way Oh_really Poland Wow snap_out_of_it just_lovely fortitude California humility patience why_do_I_put_up_with_this ingrate begs_the_question pet caution anger what_would_Jesus_do figuratively other umm_the_other_answer later how_high Norway yada_yada_yada

----

India-nigger programmers don't know how to use a hash table to collect unique words from a document. "But it's not secure without salt!"

----

For what it's worth, math skill is pretty synanomous with IQ -- the real world idea of IQ. Sure, liberals can claim other things are IQ, but nobody really believes it. Wisdom, by the way, is different from IQ.

God says... his house, and from his labour, that performeth not this promise, even thus be he shaken out, and emptied. And all the congregation said, Amen, and praised the LORD. And the people did according to this promise.

5:14 Moreover from the time that I was appointed to be their governor in the land of Judah, from the twentieth year even unto the two and thirtieth year of Artaxerxes the king, that is, twelve years, I and my brethren have not eaten the bread of the governor.

5:15 But the former governors that had been before me were chargeable unto the people, and had taken of them bread and wine, beside forty shekels of silver; yea, even their servants bare rule over the people: but so did not I, because of the fear of God.

-----

Stupid shrink! I threw him a bone saying marketeers understood persuasive writing. Hint: that's psychology.

God says... weakest uses told accompany save death disdained hogs builded disentangle stated economy familiarly regardest likeness kindled instinctive bathing abide action basically wilfully efforts unthankful Estonia faileth fer_sure duties balneum START dance starting tempt devotion Sri_Lanka forethought forenoon heads sees heightening insensibly encumbered vow readest Image creation kindness awoke turned covered pair fool do_not_disturb no_more_tears X Glory Angels congratulation countenance direction Jupiter proper combine shrunk how_high sheep raging named frame recorded tm dear disregard com rocket_science balneum excelled frenzied gladly grave's firmament figurative resist taunt imparted glorious finds bounces Venus unlearned omened desirous constrain divine lick challenged panting visions infringement holiness this builded file agonised ran justify judgements amazed uncase perturbations Grace honied Exodus East_Timor endureth suffice limitation broke delivers You_da_man altogether astonishment incorruption sensation lengthened what_part_of_God_do_you_not_understand Jesus confidentially notices KIND thread Money perils excel shoe biggot out regulate wert written infantine unworthy waterest Episcopal rich boil observe Foul believed confidently dinner Speak murmured WHAT disputing plainer remission That's_my_favorite unforgotten been their DOMINE cogitated embalmed haphazard

Psychology is unlived wisdom, in my understanding. India-nigger is going-off on a particular x86 instruction. Go write an assembler and compiler and get back to me.

You may be overestimating the sophistication of these algorithms.

At least for this training set, my algorithm rewarded the length of the essay most of all (something like 65% of the total prediction). The only other significant factors were misspellings and prevalence of certain parts of speech.

That model matched the accuracy of human graders and several commercial essay grading packages.

Students reverse-engineering comparable algorithms won't necessarily have to write well to score well.

Will this help the nearly-illiterate masses? I think so. Will it hurt those who are one standard deviation or more above the norm? I think that's a problem, too. The solution then is to use the algorithm on the first set and more individual-oriented grading methods for the other set. Effort should be spent finding those in the other set and that isn't restricted to a single course or assignment. This challenges the assumption that students are equal and deserve equal grading, but I don't really see another way forward that doesn't hurt any of the best, mediocre, or poor writers.

currently, if a student writes nonsense, there's a fairly significant chance that they will be caught and penalised. a human can detect nonsense in three minutes.

in contrast, i suspect algorithmic approaches can be gamed more easily because they don't adapt in the same way. they're not solving the hard ai problem; they're grading essays (currently) written for a human reviewer.

for example, what happens if a child learns an existing text by heart and then substitutes appropriate nouns and verbs to suit the context? say they learn "We hold these truths to be self-evident, that all men are created equal" and then, for an essay on their favourite pet, they hand in "We hold these kittens to be furry, that all kittens are created hungry". That's good grammar; it's got suitable references to the subject; it's clearly nonsense.